{ "schema": "agent-public-bench@1", "generatedAt": "2026-10-07T00:00:00.000Z", "url": "https://agent.sasid.ai/benchmarks/inference-provider-index", "study": { "slug": "inference-provider-index", "title": "Inference provider index: 27 models, 52 providers", "seoTitle": "Inference provider prices: OpenRouter vs direct", "description": "Price per million tokens for 27 models across 52 providers, the spread between them and OpenRouter’s markup over first-party prices.", "question": "For the same model, how much do inference providers differ in price, and what does a gateway such as OpenRouter add over the first-party list price?", "answer": "OpenRouter’s public API listed 265 endpoints from 52 providers for 27 models on 2026-10-06. For 10 of the 10 models with a first-party list price, OpenRouter’s per-token price was the same as the vendor’s; the cost of the gateway is the 5.5% credit-purchase fee (Standard plan, third-party-reported), so the effective markup is +5.5%. 15 of 15 closed models (Claude, GPT, Gemini) with 2 or more providers had one price across every standard-tier provider; their price differences come from named tiers (flex, priority, fast) and regional endpoints. Open-weight models differ by provider: DeepSeek V4 Flash 0423 12.6x (15 providers), DeepSeek V4 Pro 0423 11.2x (15 providers), gpt-oss-120b 6.9x (20 providers), Llama 3.3 70B Instruct 6.7x (10 providers) between the most expensive and the cheapest standard-tier provider (a calculation on reported prices). The cheapest endpoints often report lower precision or a shorter context. Latency and throughput were not in the keyless API (0 of 265 endpoints), and the gateway’s own delay is not measured: no key in the environment.", "date": "2026-10-06", "updated": "2026-10-06", "tags": [ "inference", "providers", "openrouter", "pricing", "gateway", "open-weight", "third-party-reported" ], "method": [ "One snapshot of OpenRouter’s public, keyless API on 2026-10-06: the model list and the endpoint list of 27 curated models (28 requests). Prices, context, quantization and uptime are copied as the API reported them; a field it did not return stays empty.", "Tier: read from the endpoint tag suffix. Flex, priority, fast, ultrafast and batch are named tiers; a region suffix (us, eu, europe, a cloud region) is regional; anything else is standard. This is a heuristic.", "Per-model charts show the standard tier only, with one bar per provider: its cheapest standard endpoint. The endpoint table lists every endpoint in every tier.", "Blended price = (3 × input + output) ÷ 4, a 3:1 input:output token mix. Spread = most expensive ÷ cheapest blended price among standard-tier providers. Both are calculations on reported prices.", "Markup = OpenRouter list price ÷ first-party list price − 1. First-party prices are the vendors’ published list prices as recorded in the product price table. The effective markup adds the credit-purchase fee for card purchases on the Standard plan.", "Gateway delay (time to first token, total time, billed cost per call): not measured: no key in the environment. A live harness is ready; it sends nothing without a key and has a hard spending cap." ], "caveats": [ "Every price is third-party-reported by OpenRouter’s API at 2026-10-06. Prices change often; refetch before relying on them.", "A provider listed on OpenRouter is reached through OpenRouter; its price there may differ from the price on the provider’s own site.", "The cheapest endpoint may run lower precision (fp4 or fp8) or a shorter context. Price alone does not make two endpoints equal.", "No latency or throughput figure: the keyless API returned none. A cheaper provider is not shown to be slower or faster.", "First-party prices in the product table were verified on an earlier date than the snapshot; a vendor price change in between would show as a markup.", "Comparison rows between providers name no winner: a price has no interval, so the rule for winners does not apply. The gap is stated." ], "sourceIds": [ "openrouter-api-snapshot", "openrouter-fees", "price-anthropic", "price-openai", "price-google" ], "stats": [ { "id": "provider-index-endpoints", "label": "Provider endpoints in the snapshot", "value": 265, "unit": "count", "display": "265 endpoints, 52 providers, 27 models", "n": 27 }, { "id": "provider-index-max-spread", "label": "Largest standard-tier price spread (DeepSeek V4 Flash 0423)", "value": 12.57, "unit": "ratio", "display": "12.6x (most expensive: Cloudflare; cheapest: StreamLake (fp8))", "n": 15, "note": "Calculation on reported prices, blended 3:1." }, { "id": "gateway-zero-markup-models", "label": "Models where OpenRouter’s per-token price equals the first-party list price", "value": 10, "unit": "count", "display": "10 of 10", "n": 10 }, { "id": "gateway-credit-fee", "label": "Credit-purchase fee on OpenRouter’s Standard plan (third-party-reported)", "value": 5.5, "unit": "percent", "display": "5.5% ($0.80 minimum by card)" }, { "id": "provider-index-latency-reported", "label": "Endpoints with a latency figure in the keyless API", "value": 0, "unit": "count", "display": "0 of 265", "n": 265 } ], "charts": [ { "id": "provider-index-spread", "title": "How much more the priciest provider charges than the cheapest", "subtitle": "Standard tier, blended price (3 input : 1 output), most expensive provider ÷ cheapest provider", "kind": "bar", "unit": "ratio", "yLabel": "Most expensive ÷ cheapest", "series": [ { "name": "Price spread", "points": [ { "label": "DeepSeek V4 Flash 0423", "value": 12.57, "n": 15, "highlight": true }, { "label": "DeepSeek V4 Pro 0423", "value": 11.21, "n": 15, "highlight": true }, { "label": "gpt-oss-120b", "value": 6.92, "n": 20, "highlight": true }, { "label": "Llama 3.3 70B Instruct", "value": 6.71, "n": 10, "highlight": true }, { "label": "GLM 5.3", "value": 4.69, "n": 32, "highlight": true }, { "label": "Kimi K3", "value": 1.78, "n": 19, "highlight": false }, { "label": "Llama 4 Maverick", "value": 1.69, "n": 3, "highlight": false }, { "label": "Claude Haiku 4.5", "value": 1, "n": 4, "highlight": false }, { "label": "Claude Sonnet 5", "value": 1, "n": 5, "highlight": false }, { "label": "Claude Sonnet 5.5", "value": 1, "n": 5, "highlight": false }, { "label": "Claude Opus 4.8", "value": 1, "n": 5, "highlight": false }, { "label": "Claude Opus 5", "value": 1, "n": 5, "highlight": false }, { "label": "Claude Opus 5.5", "value": 1, "n": 5, "highlight": false }, { "label": "Claude Fable 5.1", "value": 1, "n": 4, "highlight": false }, { "label": "GPT-6 Sol", "value": 1, "n": 2, "highlight": false }, { "label": "GPT-6 Luna", "value": 1, "n": 2, "highlight": false }, { "label": "GPT-6 Astra", "value": 1, "n": 2, "highlight": false }, { "label": "GPT-5.5", "value": 1, "n": 2, "highlight": false }, { "label": "Gemini 3.8 Flash", "value": 1, "n": 2, "highlight": false }, { "label": "Gemini 3.5 Flash", "value": 1, "n": 2, "highlight": false }, { "label": "Gemini 3.5 Flash Lite", "value": 1, "n": 2, "highlight": false }, { "label": "Gemini 3.1 Pro Preview", "value": 1, "n": 2, "highlight": false } ] } ], "note": "A calculation on prices reported by OpenRouter’s public API, snapshot 2026-10-06. n = providers with a standard-tier endpoint. 1x means every provider charges the same. Highlighted: a spread of 4x or more.", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "gateway-markup-vs-first-party", "title": "OpenRouter markup over the first-party list price", "subtitle": "Per-token markup, and the markup after the Standard credit-purchase fee (calculation)", "kind": "grouped-bar", "unit": "percent", "yLabel": "Markup (%)", "series": [ { "name": "Per-token markup (input)", "points": [ { "label": "Claude Haiku 4.5", "value": 0 }, { "label": "Claude Sonnet 5", "value": 0 }, { "label": "Claude Sonnet 5.5", "value": 0 }, { "label": "Claude Opus 4.8", "value": 0 }, { "label": "Claude Opus 5", "value": 0 }, { "label": "Claude Opus 5.5", "value": 0 }, { "label": "Claude Fable 5.1", "value": 0 }, { "label": "GPT-6 Luna", "value": 0 }, { "label": "Gemini 3.8 Flash", "value": 0 }, { "label": "Gemini 3.5 Flash", "value": 0 } ] }, { "name": "Per-token markup (output)", "points": [ { "label": "Claude Haiku 4.5", "value": 0 }, { "label": "Claude Sonnet 5", "value": 0 }, { "label": "Claude Sonnet 5.5", "value": 0 }, { "label": "Claude Opus 4.8", "value": 0 }, { "label": "Claude Opus 5", "value": 0 }, { "label": "Claude Opus 5.5", "value": 0 }, { "label": "Claude Fable 5.1", "value": 0 }, { "label": "GPT-6 Luna", "value": 0 }, { "label": "Gemini 3.8 Flash", "value": 0 }, { "label": "Gemini 3.5 Flash", "value": 0 } ] }, { "name": "With the 5.5% card credit fee (input)", "points": [ { "label": "Claude Haiku 4.5", "value": 5.5 }, { "label": "Claude Sonnet 5", "value": 5.5 }, { "label": "Claude Sonnet 5.5", "value": 5.5 }, { "label": "Claude Opus 4.8", "value": 5.5 }, { "label": "Claude Opus 5", "value": 5.5 }, { "label": "Claude Opus 5.5", "value": 5.5 }, { "label": "Claude Fable 5.1", "value": 5.5 }, { "label": "GPT-6 Luna", "value": 5.5 }, { "label": "Gemini 3.8 Flash", "value": 5.5 }, { "label": "Gemini 3.5 Flash", "value": 5.5 } ] } ], "note": "A calculation: OpenRouter list price ÷ first-party list price − 1, then × (1 + 5.5%) for credits bought by card on the Standard plan (purchases large enough that the minimum fee does not apply). Fees as published on 2026-10-06; first-party prices from the vendors’ list prices.", "sourceIds": [ "openrouter-api-snapshot", "openrouter-fees", "price-anthropic", "price-openai", "price-google" ] }, { "id": "gateway-vs-direct-claude-haiku-4-5", "title": "Claude Haiku 4.5: OpenRouter vs Anthropic list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 1 }, { "label": "Anthropic (first-party list price)", "value": 1 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 5 }, { "label": "Anthropic (first-party list price)", "value": 5 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Anthropic price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Claude Haiku 4.5 · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-anthropic" ] }, { "id": "gateway-vs-direct-claude-sonnet-5", "title": "Claude Sonnet 5: OpenRouter vs Anthropic list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 2 }, { "label": "Anthropic (first-party list price)", "value": 2 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 10 }, { "label": "Anthropic (first-party list price)", "value": 10 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Anthropic price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Claude Sonnet 5 · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-anthropic" ] }, { "id": "gateway-vs-direct-claude-sonnet-5-5", "title": "Claude Sonnet 5.5: OpenRouter vs Anthropic list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 2 }, { "label": "Anthropic (first-party list price)", "value": 2 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 10 }, { "label": "Anthropic (first-party list price)", "value": 10 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Anthropic price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Claude Sonnet 5.5 · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-anthropic" ] }, { "id": "gateway-vs-direct-claude-opus-4-8", "title": "Claude Opus 4.8: OpenRouter vs Anthropic list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 5 }, { "label": "Anthropic (first-party list price)", "value": 5 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 25 }, { "label": "Anthropic (first-party list price)", "value": 25 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Anthropic price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Claude Opus 4.8 · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-anthropic" ] }, { "id": "gateway-vs-direct-claude-opus-5", "title": "Claude Opus 5: OpenRouter vs Anthropic list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 5 }, { "label": "Anthropic (first-party list price)", "value": 5 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 25 }, { "label": "Anthropic (first-party list price)", "value": 25 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Anthropic price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Claude Opus 5 · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-anthropic" ] }, { "id": "gateway-vs-direct-claude-opus-5-5", "title": "Claude Opus 5.5: OpenRouter vs Anthropic list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 4 }, { "label": "Anthropic (first-party list price)", "value": 4 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 20 }, { "label": "Anthropic (first-party list price)", "value": 20 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Anthropic price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Claude Opus 5.5 · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-anthropic" ] }, { "id": "gateway-vs-direct-claude-fable-5-1", "title": "Claude Fable 5.1: OpenRouter vs Anthropic list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 10 }, { "label": "Anthropic (first-party list price)", "value": 10 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 50 }, { "label": "Anthropic (first-party list price)", "value": 50 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Anthropic price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Claude Fable 5.1 · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-anthropic" ] }, { "id": "gateway-vs-direct-gpt-6-luna", "title": "GPT-6 Luna: OpenRouter vs OpenAI list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 0.1 }, { "label": "OpenAI (first-party list price)", "value": 0.1 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 0.5 }, { "label": "OpenAI (first-party list price)", "value": 0.5 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. OpenAI price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "GPT-6 Luna · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-openai" ] }, { "id": "gateway-vs-direct-gemini-3-8-flash", "title": "Gemini 3.8 Flash: OpenRouter vs Google list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 0.75 }, { "label": "Google AI Studio (first-party list price)", "value": 0.75 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 3.75 }, { "label": "Google AI Studio (first-party list price)", "value": 3.75 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Google price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Gemini 3.8 Flash · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-google" ] }, { "id": "gateway-vs-direct-gemini-3-5-flash", "title": "Gemini 3.5 Flash: OpenRouter vs Google list price", "subtitle": "USD per million tokens, before any credit-purchase fee", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "OpenRouter", "value": 1.5 }, { "label": "Google AI Studio (first-party list price)", "value": 1.5 } ] }, { "name": "Output", "points": [ { "label": "OpenRouter", "value": 9 }, { "label": "Google AI Studio (first-party list price)", "value": 9 } ] } ], "note": "OpenRouter price reported by OpenRouter’s public API, snapshot 2026-10-06. Google price from the vendor’s published list price as recorded in the product price table. OpenRouter charges a 5.5% fee when credits are bought, not per request; the markup chart shows the price with that fee.", "factContext": "Gemini 3.5 Flash · list price, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot", "price-google" ] }, { "id": "provider-prices-claude-haiku-4-5", "title": "Claude Haiku 4.5: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Amazon Bedrock", "value": 1 }, { "label": "Anthropic", "value": 1 }, { "label": "Azure", "value": 1 }, { "label": "Google Vertex", "value": 1 } ] }, { "name": "Output", "points": [ { "label": "Amazon Bedrock", "value": 5 }, { "label": "Anthropic", "value": 5 }, { "label": "Azure", "value": 5 }, { "label": "Google Vertex", "value": 5 } ] }, { "name": "Cache read", "points": [ { "label": "Amazon Bedrock", "value": 0.1 }, { "label": "Anthropic", "value": 0.1 }, { "label": "Azure", "value": 0.1 }, { "label": "Google Vertex", "value": 0.1 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Claude Haiku 4.5 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-claude-sonnet-5", "title": "Claude Sonnet 5: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Amazon Bedrock", "value": 2 }, { "label": "Anthropic", "value": 2 }, { "label": "Azure", "value": 2 }, { "label": "Claude Platform on AWS", "value": 2 }, { "label": "Google Vertex", "value": 2 } ] }, { "name": "Output", "points": [ { "label": "Amazon Bedrock", "value": 10 }, { "label": "Anthropic", "value": 10 }, { "label": "Azure", "value": 10 }, { "label": "Claude Platform on AWS", "value": 10 }, { "label": "Google Vertex", "value": 10 } ] }, { "name": "Cache read", "points": [ { "label": "Amazon Bedrock", "value": 0.2 }, { "label": "Anthropic", "value": 0.2 }, { "label": "Azure", "value": 0.2 }, { "label": "Claude Platform on AWS", "value": 0.2 }, { "label": "Google Vertex", "value": 0.2 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Claude Sonnet 5 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-claude-sonnet-5-5", "title": "Claude Sonnet 5.5: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Amazon Bedrock", "value": 2 }, { "label": "Anthropic", "value": 2 }, { "label": "Azure", "value": 2 }, { "label": "Claude Platform on AWS", "value": 2 }, { "label": "Google Vertex", "value": 2 } ] }, { "name": "Output", "points": [ { "label": "Amazon Bedrock", "value": 10 }, { "label": "Anthropic", "value": 10 }, { "label": "Azure", "value": 10 }, { "label": "Claude Platform on AWS", "value": 10 }, { "label": "Google Vertex", "value": 10 } ] }, { "name": "Cache read", "points": [ { "label": "Amazon Bedrock", "value": 0.2 }, { "label": "Anthropic", "value": 0.2 }, { "label": "Azure", "value": 0.2 }, { "label": "Claude Platform on AWS", "value": 0.2 }, { "label": "Google Vertex", "value": 0.2 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Claude Sonnet 5.5 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-claude-opus-4-8", "title": "Claude Opus 4.8: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Amazon Bedrock", "value": 5 }, { "label": "Anthropic", "value": 5 }, { "label": "Azure", "value": 5 }, { "label": "Claude Platform on AWS", "value": 5 }, { "label": "Google Vertex", "value": 5 } ] }, { "name": "Output", "points": [ { "label": "Amazon Bedrock", "value": 25 }, { "label": "Anthropic", "value": 25 }, { "label": "Azure", "value": 25 }, { "label": "Claude Platform on AWS", "value": 25 }, { "label": "Google Vertex", "value": 25 } ] }, { "name": "Cache read", "points": [ { "label": "Amazon Bedrock", "value": 0.5 }, { "label": "Anthropic", "value": 0.5 }, { "label": "Azure", "value": 0.5 }, { "label": "Claude Platform on AWS", "value": 0.5 }, { "label": "Google Vertex", "value": 0.5 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Claude Opus 4.8 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-claude-opus-5", "title": "Claude Opus 5: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Amazon Bedrock", "value": 5 }, { "label": "Anthropic", "value": 5 }, { "label": "Azure", "value": 5 }, { "label": "Claude Platform on AWS", "value": 5 }, { "label": "Google Vertex", "value": 5 } ] }, { "name": "Output", "points": [ { "label": "Amazon Bedrock", "value": 25 }, { "label": "Anthropic", "value": 25 }, { "label": "Azure", "value": 25 }, { "label": "Claude Platform on AWS", "value": 25 }, { "label": "Google Vertex", "value": 25 } ] }, { "name": "Cache read", "points": [ { "label": "Amazon Bedrock", "value": 0.5 }, { "label": "Anthropic", "value": 0.5 }, { "label": "Azure", "value": 0.5 }, { "label": "Claude Platform on AWS", "value": 0.5 }, { "label": "Google Vertex", "value": 0.5 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Claude Opus 5 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-claude-opus-5-5", "title": "Claude Opus 5.5: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Amazon Bedrock", "value": 4 }, { "label": "Anthropic", "value": 4 }, { "label": "Azure", "value": 4 }, { "label": "Claude Platform on AWS", "value": 4 }, { "label": "Google Vertex", "value": 4 } ] }, { "name": "Output", "points": [ { "label": "Amazon Bedrock", "value": 20 }, { "label": "Anthropic", "value": 20 }, { "label": "Azure", "value": 20 }, { "label": "Claude Platform on AWS", "value": 20 }, { "label": "Google Vertex", "value": 20 } ] }, { "name": "Cache read", "points": [ { "label": "Amazon Bedrock", "value": 0.2 }, { "label": "Anthropic", "value": 0.2 }, { "label": "Azure", "value": 0.2 }, { "label": "Claude Platform on AWS", "value": 0.2 }, { "label": "Google Vertex", "value": 0.2 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Claude Opus 5.5 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-claude-fable-5-1", "title": "Claude Fable 5.1: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Amazon Bedrock", "value": 10 }, { "label": "Anthropic", "value": 10 }, { "label": "Azure", "value": 10 }, { "label": "Google Vertex", "value": 10 } ] }, { "name": "Output", "points": [ { "label": "Amazon Bedrock", "value": 50 }, { "label": "Anthropic", "value": 50 }, { "label": "Azure", "value": 50 }, { "label": "Google Vertex", "value": 50 } ] }, { "name": "Cache read", "points": [ { "label": "Amazon Bedrock", "value": 0.25 }, { "label": "Anthropic", "value": 0.25 }, { "label": "Azure", "value": 0.25 }, { "label": "Google Vertex", "value": 0.25 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Claude Fable 5.1 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gpt-6-sol", "title": "GPT-6 Sol: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Azure", "value": 2 }, { "label": "OpenAI", "value": 2 } ] }, { "name": "Output", "points": [ { "label": "Azure", "value": 10 }, { "label": "OpenAI", "value": 10 } ] }, { "name": "Cache read", "points": [ { "label": "Azure", "value": 0.2 }, { "label": "OpenAI", "value": 0.2 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "GPT-6 Sol · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gpt-6-luna", "title": "GPT-6 Luna: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Azure", "value": 0.1 }, { "label": "OpenAI", "value": 0.1 } ] }, { "name": "Output", "points": [ { "label": "Azure", "value": 0.5 }, { "label": "OpenAI", "value": 0.5 } ] }, { "name": "Cache read", "points": [ { "label": "Azure", "value": 0.01 }, { "label": "OpenAI", "value": 0.01 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "GPT-6 Luna · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gpt-6-astra", "title": "GPT-6 Astra: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Azure", "value": 10 }, { "label": "OpenAI", "value": 10 } ] }, { "name": "Output", "points": [ { "label": "Azure", "value": 50 }, { "label": "OpenAI", "value": 50 } ] }, { "name": "Cache read", "points": [ { "label": "Azure", "value": 1 }, { "label": "OpenAI", "value": 1 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "GPT-6 Astra · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gpt-5-5", "title": "GPT-5.5: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Azure", "value": 5 }, { "label": "OpenAI", "value": 5 } ] }, { "name": "Output", "points": [ { "label": "Azure", "value": 30 }, { "label": "OpenAI", "value": 30 } ] }, { "name": "Cache read", "points": [ { "label": "Azure", "value": 0.5 }, { "label": "OpenAI", "value": 0.5 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "GPT-5.5 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gpt-oss-120b", "title": "gpt-oss-120b: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "CoreWeave (fp4)", "value": 0.03 }, { "label": "DekaLLM (bf16)", "value": 0.03 }, { "label": "DeepInfra (bf16)", "value": 0.037 }, { "label": "AkashML (bf16)", "value": 0.037 }, { "label": "Mancer 2 (fp8)", "value": 0.045 }, { "label": "Crusoe (bf16)", "value": 0.05 }, { "label": "Novita (fp4)", "value": 0.05 }, { "label": "DigitalOcean", "value": 0.06 }, { "label": "Google Vertex", "value": 0.09 }, { "label": "BaseTen (fp4)", "value": 0.1 }, { "label": "Amazon Bedrock", "value": 0.15 }, { "label": "Groq", "value": 0.15 }, { "label": "Nebius (fp4)", "value": 0.15 }, { "label": "Phala", "value": 0.15 }, { "label": "SiliconFlow (fp8)", "value": 0.15 }, { "label": "Together", "value": 0.15 }, { "label": "Parasail (fp4)", "value": 0.1 }, { "label": "Mara", "value": 0.15 }, { "label": "SambaNova", "value": 0.14 }, { "label": "Cerebras (fp16)", "value": 0.35 } ] }, { "name": "Output", "points": [ { "label": "CoreWeave (fp4)", "value": 0.17 }, { "label": "DekaLLM (bf16)", "value": 0.18 }, { "label": "DeepInfra (bf16)", "value": 0.17 }, { "label": "AkashML (bf16)", "value": 0.187 }, { "label": "Mancer 2 (fp8)", "value": 0.25 }, { "label": "Crusoe (bf16)", "value": 0.25 }, { "label": "Novita (fp4)", "value": 0.25 }, { "label": "DigitalOcean", "value": 0.42 }, { "label": "Google Vertex", "value": 0.36 }, { "label": "BaseTen (fp4)", "value": 0.5 }, { "label": "Amazon Bedrock", "value": 0.6 }, { "label": "Groq", "value": 0.6 }, { "label": "Nebius (fp4)", "value": 0.6 }, { "label": "Phala", "value": 0.6 }, { "label": "SiliconFlow (fp8)", "value": 0.6 }, { "label": "Together", "value": 0.6 }, { "label": "Parasail (fp4)", "value": 0.75 }, { "label": "Mara", "value": 0.75 }, { "label": "SambaNova", "value": 0.95 }, { "label": "Cerebras (fp16)", "value": 0.75 } ] }, { "name": "Cache read", "points": [ { "label": "CoreWeave (fp4)", "value": 0.03 }, { "label": "DekaLLM (bf16)", "value": 0.03 }, { "label": "AkashML (bf16)", "value": 0.037 }, { "label": "Crusoe (bf16)", "value": 0.05 }, { "label": "DigitalOcean", "value": 0.012 }, { "label": "BaseTen (fp4)", "value": 0.1 }, { "label": "Groq", "value": 0.075 }, { "label": "SiliconFlow (fp8)", "value": 0.075 }, { "label": "Parasail (fp4)", "value": 0.055 }, { "label": "Cerebras (fp16)", "value": 0.35 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "gpt-oss-120b · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gemini-3-8-flash", "title": "Gemini 3.8 Flash: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Google AI Studio", "value": 0.75 }, { "label": "Google Vertex", "value": 0.75 } ] }, { "name": "Output", "points": [ { "label": "Google AI Studio", "value": 3.75 }, { "label": "Google Vertex", "value": 3.75 } ] }, { "name": "Cache read", "points": [ { "label": "Google AI Studio", "value": 0.075 }, { "label": "Google Vertex", "value": 0.075 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Gemini 3.8 Flash · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gemini-3-5-flash", "title": "Gemini 3.5 Flash: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Google AI Studio", "value": 1.5 }, { "label": "Google Vertex", "value": 1.5 } ] }, { "name": "Output", "points": [ { "label": "Google AI Studio", "value": 9 }, { "label": "Google Vertex", "value": 9 } ] }, { "name": "Cache read", "points": [ { "label": "Google AI Studio", "value": 0.15 }, { "label": "Google Vertex", "value": 0.15 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Gemini 3.5 Flash · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gemini-3-5-flash-lite", "title": "Gemini 3.5 Flash Lite: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Google AI Studio", "value": 0.3 }, { "label": "Google Vertex", "value": 0.3 } ] }, { "name": "Output", "points": [ { "label": "Google AI Studio", "value": 2.5 }, { "label": "Google Vertex", "value": 2.5 } ] }, { "name": "Cache read", "points": [ { "label": "Google AI Studio", "value": 0.03 }, { "label": "Google Vertex", "value": 0.03 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Gemini 3.5 Flash Lite · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-gemini-3-1-pro-preview", "title": "Gemini 3.1 Pro Preview: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Google AI Studio", "value": 2 }, { "label": "Google Vertex", "value": 2 } ] }, { "name": "Output", "points": [ { "label": "Google AI Studio", "value": 12 }, { "label": "Google Vertex", "value": 12 } ] }, { "name": "Cache read", "points": [ { "label": "Google AI Studio", "value": 0.2 }, { "label": "Google Vertex", "value": 0.2 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Gemini 3.1 Pro Preview · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-llama-4-maverick", "title": "Llama 4 Maverick: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "DigitalOcean", "value": 0.1875 }, { "label": "Novita (fp8)", "value": 0.27 }, { "label": "Parasail (fp8)", "value": 0.35 } ] }, { "name": "Output", "points": [ { "label": "DigitalOcean", "value": 0.6525 }, { "label": "Novita (fp8)", "value": 0.85 }, { "label": "Parasail (fp8)", "value": 1 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Llama 4 Maverick · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-llama-3-3-70b-instruct", "title": "Llama 3.3 70B Instruct: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "DeepInfra (fp8)", "value": 0.1 }, { "label": "Novita (bf16)", "value": 0.135 }, { "label": "AkashML (fp8)", "value": 0.2 }, { "label": "Parasail (fp8)", "value": 0.22 }, { "label": "SambaNova", "value": 0.45 }, { "label": "Groq", "value": 0.59 }, { "label": "CoreWeave (fp16)", "value": 0.71 }, { "label": "Google Vertex", "value": 0.72 }, { "label": "Cloudflare (fp8)", "value": 0.293 }, { "label": "Together", "value": 1.04 } ] }, { "name": "Output", "points": [ { "label": "DeepInfra (fp8)", "value": 0.32 }, { "label": "Novita (bf16)", "value": 0.4 }, { "label": "AkashML (fp8)", "value": 0.52 }, { "label": "Parasail (fp8)", "value": 0.5 }, { "label": "SambaNova", "value": 0.9 }, { "label": "Groq", "value": 0.79 }, { "label": "CoreWeave (fp16)", "value": 0.71 }, { "label": "Google Vertex", "value": 0.72 }, { "label": "Cloudflare (fp8)", "value": 2.253 }, { "label": "Together", "value": 1.04 } ] }, { "name": "Cache read", "points": [ { "label": "AkashML (fp8)", "value": 0.1 }, { "label": "Parasail (fp8)", "value": 0.11 }, { "label": "Groq", "value": 0.295 }, { "label": "CoreWeave (fp16)", "value": 0.71 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Llama 3.3 70B Instruct · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-deepseek-v4-pro", "title": "DeepSeek V4 Pro 0423: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "StreamLake (fp8)", "value": 0.2088 }, { "label": "GMICloud (fp8)", "value": 0.957 }, { "label": "Relace (fp4)", "value": 0.2067 }, { "label": "Parasail (fp8)", "value": 0.45 }, { "label": "DigitalOcean", "value": 1.044 }, { "label": "Cloudflare", "value": 1.15 }, { "label": "DeepInfra (fp8)", "value": 1.3 }, { "label": "Alibaba (fp8)", "value": 1.416 }, { "label": "SiliconFlow (fp8)", "value": 1.50162 }, { "label": "Novita (fp8)", "value": 1.6 }, { "label": "Venice", "value": 1.65 }, { "label": "AtlasCloud (fp4)", "value": 1.68 }, { "label": "Baidu (fp8)", "value": 1.69 }, { "label": "NextBit (fp8)", "value": 1.74 }, { "label": "Reka", "value": 0.9 } ] }, { "name": "Output", "points": [ { "label": "StreamLake (fp8)", "value": 0.4176 }, { "label": "GMICloud (fp8)", "value": 1.914 }, { "label": "Relace (fp4)", "value": 4.2 }, { "label": "Parasail (fp8)", "value": 3.48 }, { "label": "DigitalOcean", "value": 2.088 }, { "label": "Cloudflare", "value": 2.55 }, { "label": "DeepInfra (fp8)", "value": 2.6 }, { "label": "Alibaba (fp8)", "value": 2.832 }, { "label": "SiliconFlow (fp8)", "value": 3.135 }, { "label": "Novita (fp8)", "value": 3.2 }, { "label": "Venice", "value": 3.301 }, { "label": "AtlasCloud (fp4)", "value": 3.38 }, { "label": "Baidu (fp8)", "value": 3.38 }, { "label": "NextBit (fp8)", "value": 3.48 }, { "label": "Reka", "value": 9 } ] }, { "name": "Cache read", "points": [ { "label": "StreamLake (fp8)", "value": 0.0174 }, { "label": "GMICloud (fp8)", "value": 0.07975 }, { "label": "Relace (fp4)", "value": 0.21 }, { "label": "Parasail (fp8)", "value": 0.1 }, { "label": "DigitalOcean", "value": 0.2088 }, { "label": "Cloudflare", "value": 0.2 }, { "label": "DeepInfra (fp8)", "value": 0.1 }, { "label": "Alibaba (fp8)", "value": 0.118 }, { "label": "SiliconFlow (fp8)", "value": 0.135 }, { "label": "Novita (fp8)", "value": 0.135 }, { "label": "Venice", "value": 0.33 }, { "label": "AtlasCloud (fp4)", "value": 0.13 }, { "label": "Baidu (fp8)", "value": 0.14 }, { "label": "NextBit (fp8)", "value": 0.145 }, { "label": "Reka", "value": 0.18 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "DeepSeek V4 Pro 0423 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-deepseek-v4-flash", "title": "DeepSeek V4 Flash 0423: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "StreamLake (fp8)", "value": 0.042 }, { "label": "DeepInfra (fp8)", "value": 0.09 }, { "label": "GMICloud (fp8)", "value": 0.091 }, { "label": "Venice", "value": 0.0966 }, { "label": "DigitalOcean", "value": 0.098 }, { "label": "Alibaba (fp8)", "value": 0.134 }, { "label": "SiliconFlow (fp8)", "value": 0.13 }, { "label": "AtlasCloud (fp4)", "value": 0.14 }, { "label": "Baidu (fp8)", "value": 0.14 }, { "label": "Novita (fp8)", "value": 0.14 }, { "label": "Parasail (fp8)", "value": 0.14 }, { "label": "Mancer 2 (fp8)", "value": 0.19 }, { "label": "Relace (fp4)", "value": 0.012 }, { "label": "OpenInference (fp4)", "value": 0.0132 }, { "label": "Cloudflare", "value": 0.44 } ] }, { "name": "Output", "points": [ { "label": "StreamLake (fp8)", "value": 0.084 }, { "label": "DeepInfra (fp8)", "value": 0.18 }, { "label": "GMICloud (fp8)", "value": 0.182 }, { "label": "Venice", "value": 0.1925 }, { "label": "DigitalOcean", "value": 0.196 }, { "label": "Alibaba (fp8)", "value": 0.268 }, { "label": "SiliconFlow (fp8)", "value": 0.28 }, { "label": "AtlasCloud (fp4)", "value": 0.28 }, { "label": "Baidu (fp8)", "value": 0.28 }, { "label": "Novita (fp8)", "value": 0.28 }, { "label": "Parasail (fp8)", "value": 0.28 }, { "label": "Mancer 2 (fp8)", "value": 0.5 }, { "label": "Relace (fp4)", "value": 1.28 }, { "label": "OpenInference (fp4)", "value": 1.408 }, { "label": "Cloudflare", "value": 1.32 } ] }, { "name": "Cache read", "points": [ { "label": "StreamLake (fp8)", "value": 0.0084 }, { "label": "DeepInfra (fp8)", "value": 0.018 }, { "label": "GMICloud (fp8)", "value": 0.0182 }, { "label": "Venice", "value": 0.0196 }, { "label": "DigitalOcean", "value": 0.0196 }, { "label": "Alibaba (fp8)", "value": 0.0268 }, { "label": "SiliconFlow (fp8)", "value": 0.028 }, { "label": "AtlasCloud (fp4)", "value": 0.028 }, { "label": "Baidu (fp8)", "value": 0.028 }, { "label": "Novita (fp8)", "value": 0.028 }, { "label": "Parasail (fp8)", "value": 0.07 }, { "label": "Relace (fp4)", "value": 0.012 }, { "label": "OpenInference (fp4)", "value": 0.0132 }, { "label": "Cloudflare", "value": 0.014 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "DeepSeek V4 Flash 0423 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-kimi-k3", "title": "Kimi K3: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Relace (fp4)", "value": 0.83 }, { "label": "Phala", "value": 1.95 }, { "label": "Sail Research (fp4)", "value": 0.84 }, { "label": "Decart (mxfp4)", "value": 2.01 }, { "label": "InferenceNet (fp4)", "value": 0.95 }, { "label": "Wafer", "value": 0.95 }, { "label": "Morph (fp8)", "value": 1.274 }, { "label": "Makora", "value": 1.53 }, { "label": "AkashML (fp4)", "value": 1.3 }, { "label": "DigitalOcean", "value": 2.55 }, { "label": "Together", "value": 2.7 }, { "label": "DeepInfra (mxfp4)", "value": 2.85 }, { "label": "BaseTen (fp8)", "value": 3 }, { "label": "Chutes (mxfp4)", "value": 3 }, { "label": "Fireworks", "value": 3 }, { "label": "Modal (mxfp4)", "value": 3 }, { "label": "Moonshot AI (mxfp4)", "value": 3 }, { "label": "Parasail (fp4)", "value": 3 }, { "label": "Alibaba", "value": 3.45 } ] }, { "name": "Output", "points": [ { "label": "Relace (fp4)", "value": 13 }, { "label": "Phala", "value": 9.75 }, { "label": "Sail Research (fp4)", "value": 13.5 }, { "label": "Decart (mxfp4)", "value": 10.05 }, { "label": "InferenceNet (fp4)", "value": 14 }, { "label": "Wafer", "value": 14 }, { "label": "Morph (fp8)", "value": 13.296 }, { "label": "Makora", "value": 12.75 }, { "label": "AkashML (fp4)", "value": 14 }, { "label": "DigitalOcean", "value": 12.95 }, { "label": "Together", "value": 13.5 }, { "label": "DeepInfra (mxfp4)", "value": 14.25 }, { "label": "BaseTen (fp8)", "value": 15 }, { "label": "Chutes (mxfp4)", "value": 15 }, { "label": "Fireworks", "value": 15 }, { "label": "Modal (mxfp4)", "value": 15 }, { "label": "Moonshot AI (mxfp4)", "value": 15 }, { "label": "Parasail (fp4)", "value": 15 }, { "label": "Alibaba", "value": 17.25 } ] }, { "name": "Cache read", "points": [ { "label": "Relace (fp4)", "value": 0.45 }, { "label": "Phala", "value": 0.195 }, { "label": "Sail Research (fp4)", "value": 0.3 }, { "label": "Decart (mxfp4)", "value": 0.201 }, { "label": "InferenceNet (fp4)", "value": 0.31 }, { "label": "Wafer", "value": 0.4 }, { "label": "Morph (fp8)", "value": 0.278 }, { "label": "Makora", "value": 0.204 }, { "label": "AkashML (fp4)", "value": 1.3 }, { "label": "DigitalOcean", "value": 0.255 }, { "label": "Together", "value": 0.27 }, { "label": "DeepInfra (mxfp4)", "value": 0.285 }, { "label": "BaseTen (fp8)", "value": 0.3 }, { "label": "Chutes (mxfp4)", "value": 0.3 }, { "label": "Fireworks", "value": 0.3 }, { "label": "Modal (mxfp4)", "value": 0.3 }, { "label": "Moonshot AI (mxfp4)", "value": 0.3 }, { "label": "Parasail (fp4)", "value": 0.3 }, { "label": "Alibaba", "value": 0.345 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "Kimi K3 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] }, { "id": "provider-prices-glm-5-3", "title": "GLM 5.3: price per million tokens by provider", "subtitle": "Standard tier, one bar per provider (its cheapest standard endpoint); reported by OpenRouter’s public API, snapshot 2026-10-06", "kind": "grouped-bar", "unit": "usd", "yLabel": "USD per million tokens", "series": [ { "name": "Input", "points": [ { "label": "Novita (fp8)", "value": 0.42 }, { "label": "Reka", "value": 0.17 }, { "label": "Sail Research (fp8)", "value": 0.2 }, { "label": "Morph (fp8)", "value": 0.179 }, { "label": "DeepInfra (fp4)", "value": 0.5625 }, { "label": "SiliconFlow (fp8)", "value": 0.7 }, { "label": "InferenceNet", "value": 0.14 }, { "label": "Makora (fp4)", "value": 0.18 }, { "label": "AkashML (fp8)", "value": 0.19 }, { "label": "Phala", "value": 0.84 }, { "label": "Inceptron (fp4)", "value": 0.6 }, { "label": "DigitalOcean", "value": 0.91 }, { "label": "GMICloud (fp8)", "value": 0.98 }, { "label": "Alibaba", "value": 1.19 }, { "label": "Decart (fp4)", "value": 1.19 }, { "label": "Wafer", "value": 0.15 }, { "label": "Friendli", "value": 1.26 }, { "label": "AtlasCloud (fp8)", "value": 1.4 }, { "label": "Baidu (fp8)", "value": 1.4 }, { "label": "BaseTen (fp4)", "value": 1.4 }, { "label": "Cloudflare", "value": 1.4 }, { "label": "Crusoe (fp4)", "value": 1.4 }, { "label": "Fireworks", "value": 1.4 }, { "label": "Mistral (nvfp4)", "value": 1.4 }, { "label": "Modal", "value": 1.4 }, { "label": "Nebius (fp4)", "value": 1.4 }, { "label": "Parasail (fp8)", "value": 1.4 }, { "label": "PrimeIntellect", "value": 1.4 }, { "label": "Together", "value": 1.4 }, { "label": "Venice", "value": 1.4 }, { "label": "Z.AI (fp8)", "value": 1.4 }, { "label": "Relace", "value": 0.03 } ] }, { "name": "Output", "points": [ { "label": "Novita (fp8)", "value": 1.32 }, { "label": "Reka", "value": 3 }, { "label": "Sail Research (fp8)", "value": 3.4 }, { "label": "Morph (fp8)", "value": 3.553 }, { "label": "DeepInfra (fp4)", "value": 2.5 }, { "label": "SiliconFlow (fp8)", "value": 2.2 }, { "label": "InferenceNet", "value": 4.4 }, { "label": "Makora (fp4)", "value": 4.4 }, { "label": "AkashML (fp8)", "value": 4.4 }, { "label": "Phala", "value": 2.64 }, { "label": "Inceptron (fp4)", "value": 3.39 }, { "label": "DigitalOcean", "value": 2.86 }, { "label": "GMICloud (fp8)", "value": 3.08 }, { "label": "Alibaba", "value": 3.74 }, { "label": "Decart (fp4)", "value": 3.74 }, { "label": "Wafer", "value": 7 }, { "label": "Friendli", "value": 3.96 }, { "label": "AtlasCloud (fp8)", "value": 4.4 }, { "label": "Baidu (fp8)", "value": 4.4 }, { "label": "BaseTen (fp4)", "value": 4.4 }, { "label": "Cloudflare", "value": 4.4 }, { "label": "Crusoe (fp4)", "value": 4.4 }, { "label": "Fireworks", "value": 4.4 }, { "label": "Mistral (nvfp4)", "value": 4.4 }, { "label": "Modal", "value": 4.4 }, { "label": "Nebius (fp4)", "value": 4.4 }, { "label": "Parasail (fp8)", "value": 4.4 }, { "label": "PrimeIntellect", "value": 4.4 }, { "label": "Together", "value": 4.4 }, { "label": "Venice", "value": 4.4 }, { "label": "Z.AI (fp8)", "value": 4.4 }, { "label": "Relace", "value": 12 } ] }, { "name": "Cache read", "points": [ { "label": "Novita (fp8)", "value": 0.078 }, { "label": "Reka", "value": 0.169 }, { "label": "Sail Research (fp8)", "value": 0.15 }, { "label": "Morph (fp8)", "value": 0.137 }, { "label": "DeepInfra (fp4)", "value": 0.125 }, { "label": "SiliconFlow (fp8)", "value": 0.13 }, { "label": "InferenceNet", "value": 0.07 }, { "label": "Makora (fp4)", "value": 0.19 }, { "label": "AkashML (fp8)", "value": 0.19 }, { "label": "Phala", "value": 0.156 }, { "label": "Inceptron (fp4)", "value": 0.2 }, { "label": "DigitalOcean", "value": 0.169 }, { "label": "GMICloud (fp8)", "value": 0.182 }, { "label": "Alibaba", "value": 0.238 }, { "label": "Decart (fp4)", "value": 0.1955 }, { "label": "Wafer", "value": 0.14 }, { "label": "Friendli", "value": 0.234 }, { "label": "AtlasCloud (fp8)", "value": 0.26 }, { "label": "Baidu (fp8)", "value": 0.26 }, { "label": "BaseTen (fp4)", "value": 0.14 }, { "label": "Cloudflare", "value": 0.26 }, { "label": "Crusoe (fp4)", "value": 0.26 }, { "label": "Fireworks", "value": 0.26 }, { "label": "Mistral (nvfp4)", "value": 0.14 }, { "label": "Modal", "value": 0.26 }, { "label": "Parasail (fp8)", "value": 0.26 }, { "label": "PrimeIntellect", "value": 0.26 }, { "label": "Together", "value": 0.26 }, { "label": "Venice", "value": 0.26 }, { "label": "Z.AI (fp8)", "value": 0.26 }, { "label": "Relace", "value": 0.03 } ] } ], "note": "Prices reported by OpenRouter’s public API, snapshot 2026-10-06; third-party-reported, not measured by Agent. Sorted from the lowest to the highest blended price (3 input : 1 output). A parenthesis names the quantization the provider reported; lower precision or a shorter context can explain a lower price, so check the endpoint table. Flex, priority, fast and regional endpoints are left out here because they are priced differently on purpose.", "factContext": "GLM 5.3 · reported by OpenRouter’s public API, snapshot 2026-10-06", "sourceIds": [ "openrouter-api-snapshot" ] } ], "tables": [ { "id": "provider-index-models", "title": "Provider index per model (reported by OpenRouter’s public API, snapshot 2026-10-06)", "columns": [ { "key": "model", "label": "Model", "unit": "text" }, { "key": "providers", "label": "Providers", "unit": "count" }, { "key": "endpoints", "label": "Endpoints (all tiers)", "unit": "count" }, { "key": "listPrice", "label": "OpenRouter list price, in / out (USD per M)", "unit": "text" }, { "key": "cheapest", "label": "Cheapest standard provider", "unit": "text" }, { "key": "priciest", "label": "Most expensive standard provider", "unit": "text" }, { "key": "spread", "label": "Spread (blended)", "unit": "text" }, { "key": "firstParty", "label": "First-party list price, in / out", "unit": "text" } ], "rows": [ { "model": "Claude Haiku 4.5", "providers": 4, "endpoints": 8, "listPrice": "$1.00 / $5.00", "cheapest": "Amazon Bedrock: $1.00 / $5.00", "priciest": "Google Vertex: $1.00 / $5.00", "spread": "1.0x", "firstParty": "Anthropic: $1.00 / $5.00" }, { "model": "Claude Sonnet 5", "providers": 5, "endpoints": 10, "listPrice": "$2.00 / $10.00", "cheapest": "Amazon Bedrock: $2.00 / $10.00", "priciest": "Google Vertex: $2.00 / $10.00", "spread": "1.0x", "firstParty": "Anthropic: $2.00 / $10.00" }, { "model": "Claude Sonnet 5.5", "providers": 5, "endpoints": 8, "listPrice": "$2.00 / $10.00", "cheapest": "Amazon Bedrock: $2.00 / $10.00", "priciest": "Google Vertex: $2.00 / $10.00", "spread": "1.0x", "firstParty": "Anthropic: $2.00 / $10.00" }, { "model": "Claude Opus 4.8", "providers": 5, "endpoints": 11, "listPrice": "$5.00 / $25.00", "cheapest": "Amazon Bedrock: $5.00 / $25.00", "priciest": "Google Vertex: $5.00 / $25.00", "spread": "1.0x", "firstParty": "Anthropic: $5.00 / $25.00" }, { "model": "Claude Opus 5", "providers": 5, "endpoints": 11, "listPrice": "$5.00 / $25.00", "cheapest": "Amazon Bedrock: $5.00 / $25.00", "priciest": "Google Vertex: $5.00 / $25.00", "spread": "1.0x", "firstParty": "Anthropic: $5.00 / $25.00" }, { "model": "Claude Opus 5.5", "providers": 5, "endpoints": 11, "listPrice": "$4.00 / $20.00", "cheapest": "Amazon Bedrock: $4.00 / $20.00", "priciest": "Google Vertex: $4.00 / $20.00", "spread": "1.0x", "firstParty": "Anthropic: $4.00 / $20.00" }, { "model": "Claude Fable 5.1", "providers": 4, "endpoints": 4, "listPrice": "$10.00 / $50.00", "cheapest": "Amazon Bedrock: $10.00 / $50.00", "priciest": "Google Vertex: $10.00 / $50.00", "spread": "1.0x", "firstParty": "Anthropic: $10.00 / $50.00" }, { "model": "GPT-6 Sol", "providers": 3, "endpoints": 7, "listPrice": "$2.00 / $10.00", "cheapest": "Azure: $2.00 / $10.00", "priciest": "OpenAI: $2.00 / $10.00", "spread": "1.0x", "firstParty": "not in the price table" }, { "model": "GPT-6 Luna", "providers": 3, "endpoints": 7, "listPrice": "$0.10 / $0.50", "cheapest": "Azure: $0.10 / $0.50", "priciest": "OpenAI: $0.10 / $0.50", "spread": "1.0x", "firstParty": "OpenAI: $0.10 / $0.50" }, { "model": "GPT-6 Astra", "providers": 3, "endpoints": 7, "listPrice": "$10.00 / $50.00", "cheapest": "Azure: $10.00 / $50.00", "priciest": "OpenAI: $10.00 / $50.00", "spread": "1.0x", "firstParty": "not in the price table" }, { "model": "GPT-5.5", "providers": 3, "endpoints": 7, "listPrice": "$5.00 / $30.00", "cheapest": "Azure: $5.00 / $30.00", "priciest": "OpenAI: $5.00 / $30.00", "spread": "1.0x", "firstParty": "not in the price table" }, { "model": "gpt-oss-120b", "providers": 20, "endpoints": 23, "listPrice": "$0.037 / $0.17", "cheapest": "CoreWeave (fp4): $0.03 / $0.17", "priciest": "Cerebras (fp16): $0.35 / $0.75", "spread": "6.9x", "firstParty": "not in the price table" }, { "model": "Gemini 3.8 Flash", "providers": 2, "endpoints": 6, "listPrice": "$0.75 / $3.75", "cheapest": "Google AI Studio: $0.75 / $3.75", "priciest": "Google Vertex: $0.75 / $3.75", "spread": "1.0x", "firstParty": "Google: $0.75 / $3.75" }, { "model": "Gemini 3.5 Flash", "providers": 2, "endpoints": 7, "listPrice": "$1.50 / $9.00", "cheapest": "Google AI Studio: $1.50 / $9.00", "priciest": "Google Vertex: $1.50 / $9.00", "spread": "1.0x", "firstParty": "Google: $1.50 / $9.00" }, { "model": "Gemini 3.5 Flash Lite", "providers": 2, "endpoints": 8, "listPrice": "$0.30 / $2.50", "cheapest": "Google AI Studio: $0.30 / $2.50", "priciest": "Google Vertex: $0.30 / $2.50", "spread": "1.0x", "firstParty": "not in the price table" }, { "model": "Gemini 3.1 Pro Preview", "providers": 2, "endpoints": 6, "listPrice": "$2.00 / $12.00", "cheapest": "Google AI Studio: $2.00 / $12.00", "priciest": "Google Vertex: $2.00 / $12.00", "spread": "1.0x", "firstParty": "not in the price table" }, { "model": "Llama 4 Maverick", "providers": 4, "endpoints": 4, "listPrice": "$0.19 / $0.65", "cheapest": "DigitalOcean: $0.19 / $0.65", "priciest": "Parasail (fp8): $0.35 / $1.00", "spread": "1.7x", "firstParty": "not in the price table" }, { "model": "Llama 3.3 70B Instruct", "providers": 10, "endpoints": 11, "listPrice": "$0.22 / $0.50", "cheapest": "DeepInfra (fp8): $0.10 / $0.32", "priciest": "Together: $1.04 / $1.04", "spread": "6.7x", "firstParty": "not in the price table" }, { "model": "DeepSeek V4 Pro 0423", "providers": 16, "endpoints": 16, "listPrice": "$0.21 / $0.42", "cheapest": "StreamLake (fp8): $0.21 / $0.42", "priciest": "Reka: $0.90 / $9.00", "spread": "11.2x", "firstParty": "not in the price table" }, { "model": "DeepSeek V4 Flash 0423", "providers": 16, "endpoints": 16, "listPrice": "$0.012 / $1.28", "cheapest": "StreamLake (fp8): $0.042 / $0.084", "priciest": "Cloudflare: $0.44 / $1.32", "spread": "12.6x", "firstParty": "not in the price table" }, { "model": "Qwen3.8 Max (0902)", "providers": 1, "endpoints": 1, "listPrice": "$2.00 / $6.00", "cheapest": "Alibaba: $2.00 / $6.00", "priciest": "only one provider", "spread": "n/a", "firstParty": "not in the price table" }, { "model": "Qwen3.8 Flash", "providers": 1, "endpoints": 1, "listPrice": "$0.15 / $0.47", "cheapest": "Alibaba: $0.15 / $0.47", "priciest": "only one provider", "spread": "n/a", "firstParty": "not in the price table" }, { "model": "Mistral Medium 3.5", "providers": 1, "endpoints": 3, "listPrice": "$1.50 / $7.50", "cheapest": "Mistral: $1.50 / $7.50", "priciest": "only one provider", "spread": "n/a", "firstParty": "not in the price table" }, { "model": "Mistral Large 3 2512", "providers": 1, "endpoints": 2, "listPrice": "$0.50 / $1.50", "cheapest": "Mistral: $0.50 / $1.50", "priciest": "only one provider", "spread": "n/a", "firstParty": "not in the price table" }, { "model": "Grok 4.7", "providers": 1, "endpoints": 5, "listPrice": "$2.00 / $6.00", "cheapest": "xAI: $2.00 / $6.00", "priciest": "only one provider", "spread": "n/a", "firstParty": "not in the price table" }, { "model": "Kimi K3", "providers": 20, "endpoints": 24, "listPrice": "$0.95 / $14.00", "cheapest": "Relace (fp4): $0.83 / $13.00", "priciest": "Alibaba: $3.45 / $17.25", "spread": "1.8x", "firstParty": "not in the price table" }, { "model": "GLM 5.3", "providers": 32, "endpoints": 41, "listPrice": "$0.07 / $7.00", "cheapest": "Novita (fp8): $0.42 / $1.32", "priciest": "Relace: $0.03 / $12.00", "spread": "4.7x", "firstParty": "not in the price table" } ] }, { "id": "gateway-overhead-status", "title": "Gateway overhead: what is and is not measured", "columns": [ { "key": "metric", "label": "Metric", "unit": "text" }, { "key": "status", "label": "Status", "unit": "text" } ], "rows": [ { "metric": "Per-token price through OpenRouter vs first-party", "status": "Reported list prices, snapshot 2026-10-06" }, { "metric": "Credit-purchase fee", "status": "5.5% on Standard (third-party-reported, 2026-10-06)" }, { "metric": "Provider latency and throughput (last 30 min)", "status": "unknown: the keyless API returned none for 265 endpoints" }, { "metric": "Time to first token and total time through the gateway vs direct", "status": "not measured: no key in the environment. The live harness sends nothing without an OpenRouter key." }, { "metric": "Billed cost per call through the gateway", "status": "not measured: no key in the environment. The live harness sends nothing without an OpenRouter key." } ] }, { "id": "provider-index-endpoints", "title": "Every endpoint OpenRouter listed (reported by OpenRouter’s public API, snapshot 2026-10-06)", "columns": [ { "key": "model", "label": "Model", "unit": "text" }, { "key": "provider", "label": "Provider", "unit": "text" }, { "key": "tier", "label": "Tier", "unit": "text" }, { "key": "quantization", "label": "Quantization", "unit": "text" }, { "key": "context", "label": "Context (tokens)", "unit": "tokens" }, { "key": "input", "label": "Input (USD per M)", "unit": "usd" }, { "key": "output", "label": "Output (USD per M)", "unit": "usd" }, { "key": "cacheRead", "label": "Cache read (USD per M)", "unit": "usd" }, { "key": "uptime1d", "label": "Uptime, last day (%)", "unit": "percent" } ], "rows": [ { "model": "Claude Fable 5.1", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 10, "output": 50, "cacheRead": 0.25, "uptime1d": 99.36 }, { "model": "Claude Fable 5.1", "provider": "Anthropic", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 10, "output": 50, "cacheRead": 0.25, "uptime1d": 99.65 }, { "model": "Claude Fable 5.1", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 10, "output": 50, "cacheRead": 0.25, "uptime1d": null }, { "model": "Claude Fable 5.1", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 10, "output": 50, "cacheRead": 0.25, "uptime1d": 99.92 }, { "model": "Claude Haiku 4.5", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 200000, "input": 1, "output": 5, "cacheRead": 0.1, "uptime1d": 99.93 }, { "model": "Claude Haiku 4.5", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 200000, "input": 1, "output": 5, "cacheRead": 0.1, "uptime1d": 99.96 }, { "model": "Claude Haiku 4.5", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 200000, "input": 1, "output": 5, "cacheRead": 0.1, "uptime1d": 99.65 }, { "model": "Claude Haiku 4.5", "provider": "Anthropic", "tier": "standard", "quantization": "not reported", "context": 200000, "input": 1, "output": 5, "cacheRead": 0.1, "uptime1d": 99.97 }, { "model": "Claude Haiku 4.5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 200000, "input": 1.1, "output": 5.5, "cacheRead": 0.11, "uptime1d": 99.88 }, { "model": "Claude Haiku 4.5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 200000, "input": 1.1, "output": 5.5, "cacheRead": 0.11, "uptime1d": 100 }, { "model": "Claude Haiku 4.5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 200000, "input": 1.1, "output": 5.5, "cacheRead": 0.11, "uptime1d": 99.74 }, { "model": "Claude Haiku 4.5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 200000, "input": 1.1, "output": 5.5, "cacheRead": 0.11, "uptime1d": 100 }, { "model": "Claude Opus 4.8", "provider": "Claude Platform on AWS", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 99.99 }, { "model": "Claude Opus 4.8", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 98.13 }, { "model": "Claude Opus 4.8", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 94.89 }, { "model": "Claude Opus 4.8", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 99.97 }, { "model": "Claude Opus 4.8", "provider": "Anthropic", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 99.99 }, { "model": "Claude Opus 4.8", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": null }, { "model": "Claude Opus 4.8", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "Claude Opus 4.8", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": null }, { "model": "Claude Opus 4.8", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "Claude Opus 4.8", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "Claude Opus 4.8", "provider": "Anthropic", "tier": "fast", "quantization": "not reported", "context": 1000000, "input": 10, "output": 50, "cacheRead": 1, "uptime1d": 100 }, { "model": "Claude Opus 5", "provider": "Claude Platform on AWS", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 99.35 }, { "model": "Claude Opus 5", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 99.97 }, { "model": "Claude Opus 5", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 99.94 }, { "model": "Claude Opus 5", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 100 }, { "model": "Claude Opus 5", "provider": "Anthropic", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 99.86 }, { "model": "Claude Opus 5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": 99.74 }, { "model": "Claude Opus 5", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": null }, { "model": "Claude Opus 5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "Claude Opus 5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "Claude Opus 5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 5.5, "output": 27.5, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "Claude Opus 5", "provider": "Anthropic", "tier": "fast", "quantization": "not reported", "context": 1000000, "input": 10, "output": 50, "cacheRead": 1, "uptime1d": 100 }, { "model": "Claude Opus 5.5", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 4, "output": 20, "cacheRead": 0.2, "uptime1d": 97.21 }, { "model": "Claude Opus 5.5", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 4, "output": 20, "cacheRead": 0.2, "uptime1d": 99.99 }, { "model": "Claude Opus 5.5", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 4, "output": 20, "cacheRead": 0.2, "uptime1d": 99.96 }, { "model": "Claude Opus 5.5", "provider": "Claude Platform on AWS", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 4, "output": 20, "cacheRead": 0.2, "uptime1d": 99.91 }, { "model": "Claude Opus 5.5", "provider": "Anthropic", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 4, "output": 20, "cacheRead": 0.2, "uptime1d": 99.94 }, { "model": "Claude Opus 5.5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 4.4, "output": 22, "cacheRead": 0.22, "uptime1d": 99.9 }, { "model": "Claude Opus 5.5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 4.4, "output": 22, "cacheRead": 0.22, "uptime1d": 99.84 }, { "model": "Claude Opus 5.5", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 4.4, "output": 22, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "Claude Opus 5.5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 4.4, "output": 22, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "Claude Opus 5.5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 4.4, "output": 22, "cacheRead": 0.22, "uptime1d": 99.99 }, { "model": "Claude Opus 5.5", "provider": "Anthropic", "tier": "fast", "quantization": "not reported", "context": 1000000, "input": 8, "output": 40, "cacheRead": 0.4, "uptime1d": 100 }, { "model": "Claude Sonnet 5", "provider": "Claude Platform on AWS", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 100 }, { "model": "Claude Sonnet 5", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.74 }, { "model": "Claude Sonnet 5", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.99 }, { "model": "Claude Sonnet 5", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.99 }, { "model": "Claude Sonnet 5", "provider": "Anthropic", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.98 }, { "model": "Claude Sonnet 5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "Claude Sonnet 5", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": null }, { "model": "Claude Sonnet 5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "Claude Sonnet 5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 99.94 }, { "model": "Claude Sonnet 5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "Claude Sonnet 5.5", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.99 }, { "model": "Claude Sonnet 5.5", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.98 }, { "model": "Claude Sonnet 5.5", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.99 }, { "model": "Claude Sonnet 5.5", "provider": "Claude Platform on AWS", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.98 }, { "model": "Claude Sonnet 5.5", "provider": "Anthropic", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.98 }, { "model": "Claude Sonnet 5.5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 99.99 }, { "model": "Claude Sonnet 5.5", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "Claude Sonnet 5.5", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1000000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Relace", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.012, "output": 1.28, "cacheRead": 0.012, "uptime1d": 99.92 }, { "model": "DeepSeek V4 Flash 0423", "provider": "OpenInference", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.0132, "output": 1.408, "cacheRead": 0.0132, "uptime1d": 99.08 }, { "model": "DeepSeek V4 Flash 0423", "provider": "StreamLake", "tier": "standard", "quantization": "fp8", "context": 1024000, "input": 0.042, "output": 0.084, "cacheRead": 0.0084, "uptime1d": 99.12 }, { "model": "DeepSeek V4 Flash 0423", "provider": "DeepInfra", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.09, "output": 0.18, "cacheRead": 0.018, "uptime1d": 99.67 }, { "model": "DeepSeek V4 Flash 0423", "provider": "GMICloud", "tier": "standard", "quantization": "fp8", "context": 1048575, "input": 0.091, "output": 0.182, "cacheRead": 0.0182, "uptime1d": 99.98 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Venice", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 0.0966, "output": 0.1925, "cacheRead": 0.0196, "uptime1d": 96.14 }, { "model": "DeepSeek V4 Flash 0423", "provider": "DigitalOcean", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.098, "output": 0.196, "cacheRead": 0.0196, "uptime1d": 99.75 }, { "model": "DeepSeek V4 Flash 0423", "provider": "SiliconFlow", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.13, "output": 0.28, "cacheRead": 0.028, "uptime1d": 99.54 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Alibaba", "tier": "standard", "quantization": "fp8", "context": 1000000, "input": 0.134, "output": 0.268, "cacheRead": 0.0268, "uptime1d": 99.21 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Baidu", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.14, "output": 0.28, "cacheRead": 0.028, "uptime1d": 99.41 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Novita", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.14, "output": 0.28, "cacheRead": 0.028, "uptime1d": 99.87 }, { "model": "DeepSeek V4 Flash 0423", "provider": "AtlasCloud", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.14, "output": 0.28, "cacheRead": 0.028, "uptime1d": 99.09 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Parasail", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.14, "output": 0.28, "cacheRead": 0.07, "uptime1d": 99.68 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Mancer 2", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.19, "output": 0.5, "cacheRead": null, "uptime1d": 94.91 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 0.21, "output": 0.56, "cacheRead": 0.031, "uptime1d": 95.28 }, { "model": "DeepSeek V4 Flash 0423", "provider": "Cloudflare", "tier": "standard", "quantization": "not reported", "context": 384000, "input": 0.44, "output": 1.32, "cacheRead": 0.014, "uptime1d": 98.13 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Relace", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.2067, "output": 4.2, "cacheRead": 0.21, "uptime1d": 99.9 }, { "model": "DeepSeek V4 Pro 0423", "provider": "StreamLake", "tier": "standard", "quantization": "fp8", "context": 1024000, "input": 0.2088, "output": 0.4176, "cacheRead": 0.0174, "uptime1d": 98.67 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Parasail", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.45, "output": 3.48, "cacheRead": 0.1, "uptime1d": 98.3 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Reka", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.9, "output": 9, "cacheRead": 0.18, "uptime1d": 98.83 }, { "model": "DeepSeek V4 Pro 0423", "provider": "GMICloud", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.957, "output": 1.914, "cacheRead": 0.07975, "uptime1d": 97.05 }, { "model": "DeepSeek V4 Pro 0423", "provider": "DigitalOcean", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.044, "output": 2.088, "cacheRead": 0.2088, "uptime1d": 99.54 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Cloudflare", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.15, "output": 2.55, "cacheRead": 0.2, "uptime1d": 98.29 }, { "model": "DeepSeek V4 Pro 0423", "provider": "DeepInfra", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.3, "output": 2.6, "cacheRead": 0.1, "uptime1d": 99.81 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Alibaba", "tier": "standard", "quantization": "fp8", "context": 1000000, "input": 1.416, "output": 2.832, "cacheRead": 0.118, "uptime1d": 90.11 }, { "model": "DeepSeek V4 Pro 0423", "provider": "SiliconFlow", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.50162, "output": 3.135, "cacheRead": 0.135, "uptime1d": 99.21 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Novita", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.6, "output": 3.2, "cacheRead": 0.135, "uptime1d": 99.94 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Venice", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 1.65, "output": 3.301, "cacheRead": 0.33, "uptime1d": 97.05 }, { "model": "DeepSeek V4 Pro 0423", "provider": "AtlasCloud", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 1.68, "output": 3.38, "cacheRead": 0.13, "uptime1d": 99.01 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Baidu", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.69, "output": 3.38, "cacheRead": 0.14, "uptime1d": 99.96 }, { "model": "DeepSeek V4 Pro 0423", "provider": "NextBit", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.74, "output": 3.48, "cacheRead": 0.145, "uptime1d": 98.71 }, { "model": "DeepSeek V4 Pro 0423", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 1.91, "output": 3.83, "cacheRead": 0.16, "uptime1d": 99.2 }, { "model": "Gemini 3.1 Pro Preview", "provider": "Google Vertex", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 1, "output": 6, "cacheRead": 0.1, "uptime1d": 93.19 }, { "model": "Gemini 3.1 Pro Preview", "provider": "Google AI Studio", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 1, "output": 6, "cacheRead": 0.1, "uptime1d": 99.97 }, { "model": "Gemini 3.1 Pro Preview", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 2, "output": 12, "cacheRead": 0.2, "uptime1d": 98 }, { "model": "Gemini 3.1 Pro Preview", "provider": "Google AI Studio", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 2, "output": 12, "cacheRead": 0.2, "uptime1d": 99.81 }, { "model": "Gemini 3.1 Pro Preview", "provider": "Google Vertex", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 3.6, "output": 21.6, "cacheRead": 0.36, "uptime1d": 99.87 }, { "model": "Gemini 3.1 Pro Preview", "provider": "Google AI Studio", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 3.6, "output": 21.6, "cacheRead": 0.36, "uptime1d": 98.9 }, { "model": "Gemini 3.5 Flash", "provider": "Google Vertex", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 0.75, "output": 4.5, "cacheRead": 0.075, "uptime1d": 98.79 }, { "model": "Gemini 3.5 Flash", "provider": "Google AI Studio", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 0.75, "output": 4.5, "cacheRead": 0.075, "uptime1d": 99.96 }, { "model": "Gemini 3.5 Flash", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.5, "output": 9, "cacheRead": 0.15, "uptime1d": 98.72 }, { "model": "Gemini 3.5 Flash", "provider": "Google AI Studio", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.5, "output": 9, "cacheRead": 0.15, "uptime1d": 99.9 }, { "model": "Gemini 3.5 Flash", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 1.65, "output": 9.9, "cacheRead": 0.165, "uptime1d": null }, { "model": "Gemini 3.5 Flash", "provider": "Google Vertex", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 2.7, "output": 16.2, "cacheRead": 0.27, "uptime1d": 99.9 }, { "model": "Gemini 3.5 Flash", "provider": "Google AI Studio", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 2.7, "output": 16.2, "cacheRead": 0.27, "uptime1d": 99.93 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google Vertex", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 0.15, "output": 1.25, "cacheRead": 0.015, "uptime1d": 99.99 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google AI Studio", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 0.15, "output": 1.25, "cacheRead": 0.015, "uptime1d": 99.98 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google AI Studio", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.3, "output": 2.5, "cacheRead": 0.03, "uptime1d": 99.95 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.3, "output": 2.5, "cacheRead": 0.03, "uptime1d": 99.95 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 0.33, "output": 2.75, "cacheRead": 0.033, "uptime1d": 100 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 0.33, "output": 2.75, "cacheRead": 0.033, "uptime1d": 99.86 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google Vertex", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 0.54, "output": 4.5, "cacheRead": 0.054, "uptime1d": 99.95 }, { "model": "Gemini 3.5 Flash Lite", "provider": "Google AI Studio", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 0.54, "output": 4.5, "cacheRead": 0.054, "uptime1d": 99.96 }, { "model": "Gemini 3.8 Flash", "provider": "Google AI Studio", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 0.375, "output": 1.875, "cacheRead": 0.0375, "uptime1d": 99.93 }, { "model": "Gemini 3.8 Flash", "provider": "Google Vertex", "tier": "flex", "quantization": "not reported", "context": 1048576, "input": 0.375, "output": 1.875, "cacheRead": 0.0375, "uptime1d": 99.44 }, { "model": "Gemini 3.8 Flash", "provider": "Google AI Studio", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.75, "output": 3.75, "cacheRead": 0.075, "uptime1d": 99.83 }, { "model": "Gemini 3.8 Flash", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.75, "output": 3.75, "cacheRead": 0.075, "uptime1d": 97.49 }, { "model": "Gemini 3.8 Flash", "provider": "Google AI Studio", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 1.35, "output": 6.75, "cacheRead": 0.135, "uptime1d": 99.71 }, { "model": "Gemini 3.8 Flash", "provider": "Google Vertex", "tier": "priority", "quantization": "not reported", "context": 1048576, "input": 1.35, "output": 6.75, "cacheRead": 0.135, "uptime1d": 99.75 }, { "model": "GLM 5.3", "provider": "Relace", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.03, "output": 12, "cacheRead": 0.03, "uptime1d": 99.96 }, { "model": "GLM 5.3", "provider": "Wafer", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 0.07, "output": 7, "cacheRead": 0.065, "uptime1d": 99.8 }, { "model": "GLM 5.3", "provider": "InferenceNet", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.14, "output": 4.4, "cacheRead": 0.07, "uptime1d": 99.72 }, { "model": "GLM 5.3", "provider": "Wafer", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.15, "output": 7, "cacheRead": 0.14, "uptime1d": 99.6 }, { "model": "GLM 5.3", "provider": "Reka", "tier": "standard", "quantization": "not reported", "context": 262144, "input": 0.17, "output": 3, "cacheRead": 0.169, "uptime1d": 99.76 }, { "model": "GLM 5.3", "provider": "Morph", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.179, "output": 3.553, "cacheRead": 0.137, "uptime1d": 98.66 }, { "model": "GLM 5.3", "provider": "Makora", "tier": "standard", "quantization": "fp4", "context": 980000, "input": 0.18, "output": 4.4, "cacheRead": 0.19, "uptime1d": 96.34 }, { "model": "GLM 5.3", "provider": "AkashML", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.19, "output": 4.4, "cacheRead": 0.19, "uptime1d": 99.93 }, { "model": "GLM 5.3", "provider": "Sail Research", "tier": "regional", "quantization": "fp8", "context": 1048576, "input": 0.2, "output": 3.4, "cacheRead": 0.15, "uptime1d": 99.48 }, { "model": "GLM 5.3", "provider": "Sail Research", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.2, "output": 3.4, "cacheRead": 0.15, "uptime1d": 99.09 }, { "model": "GLM 5.3", "provider": "Novita", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.42, "output": 1.32, "cacheRead": 0.078, "uptime1d": 97.23 }, { "model": "GLM 5.3", "provider": "DeepInfra", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.5625, "output": 2.5, "cacheRead": 0.125, "uptime1d": 97.25 }, { "model": "GLM 5.3", "provider": "Inceptron", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.6, "output": 3.39, "cacheRead": 0.2, "uptime1d": 98.29 }, { "model": "GLM 5.3", "provider": "SiliconFlow", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.7, "output": 2.2, "cacheRead": 0.13, "uptime1d": 99.84 }, { "model": "GLM 5.3", "provider": "Phala", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.84, "output": 2.64, "cacheRead": 0.156, "uptime1d": 99.16 }, { "model": "GLM 5.3", "provider": "DigitalOcean", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.91, "output": 2.86, "cacheRead": 0.169, "uptime1d": 99.8 }, { "model": "GLM 5.3", "provider": "GMICloud", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.98, "output": 3.08, "cacheRead": 0.182, "uptime1d": 98.75 }, { "model": "GLM 5.3", "provider": "Alibaba", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 1.19, "output": 3.74, "cacheRead": 0.238, "uptime1d": 99.95 }, { "model": "GLM 5.3", "provider": "Decart", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 1.19, "output": 3.74, "cacheRead": 0.1955, "uptime1d": 99.99 }, { "model": "GLM 5.3", "provider": "Friendli", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.26, "output": 3.96, "cacheRead": 0.234, "uptime1d": 99.94 }, { "model": "GLM 5.3", "provider": "Mistral", "tier": "standard", "quantization": "nvfp4", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.14, "uptime1d": 99.73 }, { "model": "GLM 5.3", "provider": "Baidu", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 99.89 }, { "model": "GLM 5.3", "provider": "BaseTen", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.14, "uptime1d": 97.75 }, { "model": "GLM 5.3", "provider": "Mistral", "tier": "standard", "quantization": "nvfp4", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.14, "uptime1d": 99.34 }, { "model": "GLM 5.3", "provider": "Nebius", "tier": "standard", "quantization": "fp4", "context": 1024000, "input": 1.4, "output": 4.4, "cacheRead": null, "uptime1d": 96.05 }, { "model": "GLM 5.3", "provider": "Crusoe", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 98.48 }, { "model": "GLM 5.3", "provider": "PrimeIntellect", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 99.47 }, { "model": "GLM 5.3", "provider": "Venice", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 98.63 }, { "model": "GLM 5.3", "provider": "Together", "tier": "standard", "quantization": "not reported", "context": 1048575, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 96.37 }, { "model": "GLM 5.3", "provider": "Parasail", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 99.7 }, { "model": "GLM 5.3", "provider": "Modal", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 98.13 }, { "model": "GLM 5.3", "provider": "BaseTen", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.14, "uptime1d": 95.1 }, { "model": "GLM 5.3", "provider": "Fireworks", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 99.5 }, { "model": "GLM 5.3", "provider": "Cloudflare", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 98.18 }, { "model": "GLM 5.3", "provider": "AtlasCloud", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 99.94 }, { "model": "GLM 5.3", "provider": "Z.AI", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.4, "output": 4.4, "cacheRead": 0.26, "uptime1d": 99.84 }, { "model": "GLM 5.3", "provider": "Mistral", "tier": "standard", "quantization": "nvfp4", "context": 1048576, "input": 1.54, "output": 4.84, "cacheRead": 0.154, "uptime1d": 99.82 }, { "model": "GLM 5.3", "provider": "Fireworks", "tier": "fast", "quantization": "not reported", "context": 1048576, "input": 2.1, "output": 6.6, "cacheRead": 0.39, "uptime1d": 99.59 }, { "model": "GLM 5.3", "provider": "BaseTen", "tier": "fast", "quantization": "fp8", "context": 1048576, "input": 2.1, "output": 6.6, "cacheRead": 0.21, "uptime1d": 99.65 }, { "model": "GLM 5.3", "provider": "BaseTen", "tier": "fast", "quantization": "fp8", "context": 1048576, "input": 2.1, "output": 6.6, "cacheRead": 0.21, "uptime1d": 99.21 }, { "model": "GLM 5.3", "provider": "Alibaba", "tier": "fast", "quantization": "not reported", "context": 1000000, "input": 2.8, "output": 8.8, "cacheRead": 0.56, "uptime1d": 100 }, { "model": "GPT-5.5", "provider": "OpenAI", "tier": "flex", "quantization": "not reported", "context": 1050000, "input": 2.5, "output": 15, "cacheRead": 0.25, "uptime1d": 100 }, { "model": "GPT-5.5", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 5, "output": 30, "cacheRead": 0.5, "uptime1d": 99.96 }, { "model": "GPT-5.5", "provider": "OpenAI", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 5, "output": 30, "cacheRead": 0.5, "uptime1d": 99.99 }, { "model": "GPT-5.5", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 5.5, "output": 33, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "GPT-5.5", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 5.5, "output": 33, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "GPT-5.5", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 5.5, "output": 33, "cacheRead": 0.55, "uptime1d": null }, { "model": "GPT-5.5", "provider": "OpenAI", "tier": "fast", "quantization": "not reported", "context": 1050000, "input": 12.5, "output": 75, "cacheRead": 1.25, "uptime1d": 100 }, { "model": "GPT-6 Astra", "provider": "OpenAI", "tier": "flex", "quantization": "not reported", "context": 1050000, "input": 5, "output": 25, "cacheRead": 0.5, "uptime1d": 100 }, { "model": "GPT-6 Astra", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 10, "output": 50, "cacheRead": 1, "uptime1d": 99.97 }, { "model": "GPT-6 Astra", "provider": "OpenAI", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 10, "output": 50, "cacheRead": 1, "uptime1d": 99.99 }, { "model": "GPT-6 Astra", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 11, "output": 55, "cacheRead": 1.1, "uptime1d": null }, { "model": "GPT-6 Astra", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 11, "output": 55, "cacheRead": 1.1, "uptime1d": 100 }, { "model": "GPT-6 Astra", "provider": "OpenAI", "tier": "fast", "quantization": "not reported", "context": 1050000, "input": 20, "output": 100, "cacheRead": 2, "uptime1d": 99.97 }, { "model": "GPT-6 Astra", "provider": "OpenAI", "tier": "ultrafast", "quantization": "not reported", "context": 1050000, "input": 60, "output": 300, "cacheRead": 6, "uptime1d": 100 }, { "model": "GPT-6 Luna", "provider": "OpenAI", "tier": "flex", "quantization": "not reported", "context": 1050000, "input": 0.05, "output": 0.25, "cacheRead": 0.005, "uptime1d": 97.74 }, { "model": "GPT-6 Luna", "provider": "OpenAI", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 0.1, "output": 0.5, "cacheRead": 0.01, "uptime1d": 99.99 }, { "model": "GPT-6 Luna", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 0.1, "output": 0.5, "cacheRead": 0.01, "uptime1d": 99.92 }, { "model": "GPT-6 Luna", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 0.11, "output": 0.55, "cacheRead": 0.011, "uptime1d": 99.99 }, { "model": "GPT-6 Luna", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 0.11, "output": 0.55, "cacheRead": 0.011, "uptime1d": 99.98 }, { "model": "GPT-6 Luna", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 0.11, "output": 0.55, "cacheRead": 0.011, "uptime1d": 98.8 }, { "model": "GPT-6 Luna", "provider": "OpenAI", "tier": "fast", "quantization": "not reported", "context": 1050000, "input": 0.2, "output": 1, "cacheRead": 0.02, "uptime1d": 99.99 }, { "model": "GPT-6 Sol", "provider": "OpenAI", "tier": "flex", "quantization": "not reported", "context": 1050000, "input": 1, "output": 5, "cacheRead": 0.1, "uptime1d": 99.99 }, { "model": "GPT-6 Sol", "provider": "OpenAI", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 100 }, { "model": "GPT-6 Sol", "provider": "Azure", "tier": "standard", "quantization": "not reported", "context": 1050000, "input": 2, "output": 10, "cacheRead": 0.2, "uptime1d": 99.99 }, { "model": "GPT-6 Sol", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 99.98 }, { "model": "GPT-6 Sol", "provider": "Azure", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "GPT-6 Sol", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1050000, "input": 2.2, "output": 11, "cacheRead": 0.22, "uptime1d": 100 }, { "model": "GPT-6 Sol", "provider": "OpenAI", "tier": "fast", "quantization": "not reported", "context": 1050000, "input": 4, "output": 20, "cacheRead": 0.4, "uptime1d": 99.99 }, { "model": "gpt-oss-120b", "provider": "CoreWeave", "tier": "standard", "quantization": "fp4", "context": 131072, "input": 0.03, "output": 0.17, "cacheRead": 0.03, "uptime1d": 98.78 }, { "model": "gpt-oss-120b", "provider": "DekaLLM", "tier": "standard", "quantization": "bf16", "context": 131072, "input": 0.03, "output": 0.18, "cacheRead": 0.03, "uptime1d": 99.57 }, { "model": "gpt-oss-120b", "provider": "DeepInfra", "tier": "standard", "quantization": "bf16", "context": 131072, "input": 0.037, "output": 0.17, "cacheRead": null, "uptime1d": 98.93 }, { "model": "gpt-oss-120b", "provider": "AkashML", "tier": "standard", "quantization": "bf16", "context": 131072, "input": 0.037, "output": 0.187, "cacheRead": 0.037, "uptime1d": 99.98 }, { "model": "gpt-oss-120b", "provider": "Mancer 2", "tier": "standard", "quantization": "fp8", "context": 131072, "input": 0.045, "output": 0.25, "cacheRead": null, "uptime1d": 98.58 }, { "model": "gpt-oss-120b", "provider": "Crusoe", "tier": "standard", "quantization": "bf16", "context": 131072, "input": 0.05, "output": 0.25, "cacheRead": 0.05, "uptime1d": 99.98 }, { "model": "gpt-oss-120b", "provider": "Novita", "tier": "standard", "quantization": "fp4", "context": 131072, "input": 0.05, "output": 0.25, "cacheRead": null, "uptime1d": 98.57 }, { "model": "gpt-oss-120b", "provider": "DigitalOcean", "tier": "standard", "quantization": "not reported", "context": 128000, "input": 0.06, "output": 0.42, "cacheRead": 0.012, "uptime1d": 99.98 }, { "model": "gpt-oss-120b", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.09, "output": 0.36, "cacheRead": null, "uptime1d": 66.93 }, { "model": "gpt-oss-120b", "provider": "BaseTen", "tier": "standard", "quantization": "fp4", "context": 128072, "input": 0.1, "output": 0.5, "cacheRead": 0.1, "uptime1d": 99.96 }, { "model": "gpt-oss-120b", "provider": "BaseTen", "tier": "standard", "quantization": "fp4", "context": 128072, "input": 0.1, "output": 0.5, "cacheRead": 0.1, "uptime1d": 99.97 }, { "model": "gpt-oss-120b", "provider": "Parasail", "tier": "standard", "quantization": "fp4", "context": 131072, "input": 0.1, "output": 0.75, "cacheRead": 0.055, "uptime1d": 99.88 }, { "model": "gpt-oss-120b", "provider": "SambaNova", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.14, "output": 0.95, "cacheRead": null, "uptime1d": 99.68 }, { "model": "gpt-oss-120b", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": null, "uptime1d": 99.97 }, { "model": "gpt-oss-120b", "provider": "Nebius", "tier": "standard", "quantization": "fp4", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": null, "uptime1d": 97.3 }, { "model": "gpt-oss-120b", "provider": "Amazon Bedrock", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": null, "uptime1d": 99.46 }, { "model": "gpt-oss-120b", "provider": "DeepInfra", "tier": "standard", "quantization": "bf16", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": null, "uptime1d": 99.98 }, { "model": "gpt-oss-120b", "provider": "SiliconFlow", "tier": "standard", "quantization": "fp8", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": 0.075, "uptime1d": 83.35 }, { "model": "gpt-oss-120b", "provider": "Phala", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": null, "uptime1d": 99.27 }, { "model": "gpt-oss-120b", "provider": "Together", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": null, "uptime1d": 87.17 }, { "model": "gpt-oss-120b", "provider": "Groq", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.15, "output": 0.6, "cacheRead": 0.075, "uptime1d": 99.34 }, { "model": "gpt-oss-120b", "provider": "Mara", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.15, "output": 0.75, "cacheRead": null, "uptime1d": 96.56 }, { "model": "gpt-oss-120b", "provider": "Cerebras", "tier": "standard", "quantization": "fp16", "context": 131072, "input": 0.35, "output": 0.75, "cacheRead": 0.35, "uptime1d": 99.98 }, { "model": "Grok 4.7", "provider": "xAI", "tier": "standard", "quantization": "not reported", "context": 500000, "input": 2, "output": 6, "cacheRead": 0.5, "uptime1d": 99.46 }, { "model": "Grok 4.7", "provider": "xAI", "tier": "standard", "quantization": "not reported", "context": 500000, "input": 2, "output": 6, "cacheRead": 0.5, "uptime1d": 99.32 }, { "model": "Grok 4.7", "provider": "xAI", "tier": "regional", "quantization": "not reported", "context": 500000, "input": 2.2, "output": 6.6, "cacheRead": 0.55, "uptime1d": 100 }, { "model": "Grok 4.7", "provider": "xAI", "tier": "priority", "quantization": "not reported", "context": 500000, "input": 4, "output": 12, "cacheRead": 1, "uptime1d": 99.55 }, { "model": "Grok 4.7", "provider": "xAI", "tier": "priority", "quantization": "not reported", "context": 500000, "input": 4, "output": 12, "cacheRead": 1, "uptime1d": 99.38 }, { "model": "Kimi K3", "provider": "Relace", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.83, "output": 13, "cacheRead": 0.45, "uptime1d": 99.66 }, { "model": "Kimi K3", "provider": "Sail Research", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.84, "output": 13.5, "cacheRead": 0.3, "uptime1d": 99.9 }, { "model": "Kimi K3", "provider": "InferenceNet", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 0.95, "output": 14, "cacheRead": 0.31, "uptime1d": 99.9 }, { "model": "Kimi K3", "provider": "Wafer", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 0.95, "output": 14, "cacheRead": 0.4, "uptime1d": 99.48 }, { "model": "Kimi K3", "provider": "Morph", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 1.274, "output": 13.296, "cacheRead": 0.278, "uptime1d": 99.56 }, { "model": "Kimi K3", "provider": "AkashML", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 1.3, "output": 14, "cacheRead": 1.3, "uptime1d": 98.42 }, { "model": "Kimi K3", "provider": "Makora", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.53, "output": 12.75, "cacheRead": 0.204, "uptime1d": 96.99 }, { "model": "Kimi K3", "provider": "Phala", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 1.95, "output": 9.75, "cacheRead": 0.195, "uptime1d": 97.01 }, { "model": "Kimi K3", "provider": "Decart", "tier": "standard", "quantization": "mxfp4", "context": 1048576, "input": 2.01, "output": 10.05, "cacheRead": 0.201, "uptime1d": 86.93 }, { "model": "Kimi K3", "provider": "DigitalOcean", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 2.55, "output": 12.95, "cacheRead": 0.255, "uptime1d": 99.92 }, { "model": "Kimi K3", "provider": "Together", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 2.7, "output": 13.5, "cacheRead": 0.27, "uptime1d": 99.32 }, { "model": "Kimi K3", "provider": "Wafer", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 2.8, "output": 14, "cacheRead": 0.3, "uptime1d": 98.97 }, { "model": "Kimi K3", "provider": "DeepInfra", "tier": "standard", "quantization": "mxfp4", "context": 1048576, "input": 2.85, "output": 14.25, "cacheRead": 0.285, "uptime1d": 99.36 }, { "model": "Kimi K3", "provider": "Amazon Bedrock", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 3, "output": 15, "cacheRead": 0.3, "uptime1d": 93.2 }, { "model": "Kimi K3", "provider": "Chutes", "tier": "standard", "quantization": "mxfp4", "context": 1048576, "input": 3, "output": 15, "cacheRead": 0.3, "uptime1d": 97.22 }, { "model": "Kimi K3", "provider": "Parasail", "tier": "standard", "quantization": "fp4", "context": 1048576, "input": 3, "output": 15, "cacheRead": 0.3, "uptime1d": 98.44 }, { "model": "Kimi K3", "provider": "Modal", "tier": "standard", "quantization": "mxfp4", "context": 1048576, "input": 3, "output": 15, "cacheRead": 0.3, "uptime1d": 98.25 }, { "model": "Kimi K3", "provider": "Fireworks", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 3, "output": 15, "cacheRead": 0.3, "uptime1d": 99.32 }, { "model": "Kimi K3", "provider": "BaseTen", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 3, "output": 15, "cacheRead": 0.3, "uptime1d": 98.02 }, { "model": "Kimi K3", "provider": "Moonshot AI", "tier": "standard", "quantization": "mxfp4", "context": 1048576, "input": 3, "output": 15, "cacheRead": 0.3, "uptime1d": 99.99 }, { "model": "Kimi K3", "provider": "Alibaba", "tier": "standard", "quantization": "not reported", "context": 1048576, "input": 3.45, "output": 17.25, "cacheRead": 0.345, "uptime1d": 99.19 }, { "model": "Kimi K3", "provider": "InferenceNet", "tier": "fast", "quantization": "fp4", "context": 250000, "input": 3.5, "output": 15, "cacheRead": 0.45, "uptime1d": 99.8 }, { "model": "Kimi K3", "provider": "Fireworks", "tier": "regional", "quantization": "not reported", "context": 1048576, "input": 4.5, "output": 22.5, "cacheRead": 0.45, "uptime1d": 98.97 }, { "model": "Kimi K3", "provider": "Fireworks", "tier": "fast", "quantization": "not reported", "context": 1048576, "input": 4.5, "output": 22.5, "cacheRead": 0.45, "uptime1d": 96.99 }, { "model": "Llama 3.3 70B Instruct", "provider": "DeepInfra", "tier": "standard", "quantization": "fp8", "context": 131072, "input": 0.1, "output": 0.32, "cacheRead": null, "uptime1d": 98.1 }, { "model": "Llama 3.3 70B Instruct", "provider": "Novita", "tier": "standard", "quantization": "bf16", "context": 12288, "input": 0.135, "output": 0.4, "cacheRead": null, "uptime1d": 98.08 }, { "model": "Llama 3.3 70B Instruct", "provider": "AkashML", "tier": "standard", "quantization": "fp8", "context": 131072, "input": 0.2, "output": 0.52, "cacheRead": 0.1, "uptime1d": 99.27 }, { "model": "Llama 3.3 70B Instruct", "provider": "Parasail", "tier": "standard", "quantization": "fp8", "context": 131072, "input": 0.22, "output": 0.5, "cacheRead": 0.11, "uptime1d": 99.65 }, { "model": "Llama 3.3 70B Instruct", "provider": "Cloudflare", "tier": "standard", "quantization": "fp8", "context": 24000, "input": 0.293, "output": 2.253, "cacheRead": null, "uptime1d": 98.85 }, { "model": "Llama 3.3 70B Instruct", "provider": "SambaNova", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.45, "output": 0.9, "cacheRead": null, "uptime1d": 98.56 }, { "model": "Llama 3.3 70B Instruct", "provider": "Groq", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 0.59, "output": 0.79, "cacheRead": 0.295, "uptime1d": 99.85 }, { "model": "Llama 3.3 70B Instruct", "provider": "CoreWeave", "tier": "standard", "quantization": "fp16", "context": 128000, "input": 0.71, "output": 0.71, "cacheRead": 0.71, "uptime1d": 97.72 }, { "model": "Llama 3.3 70B Instruct", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 128000, "input": 0.72, "output": 0.72, "cacheRead": null, "uptime1d": null }, { "model": "Llama 3.3 70B Instruct", "provider": "Google Vertex", "tier": "standard", "quantization": "not reported", "context": 128000, "input": 0.72, "output": 0.72, "cacheRead": null, "uptime1d": null }, { "model": "Llama 3.3 70B Instruct", "provider": "Together", "tier": "standard", "quantization": "not reported", "context": 131072, "input": 1.04, "output": 1.04, "cacheRead": null, "uptime1d": 93.22 }, { "model": "Llama 4 Maverick", "provider": "DigitalOcean", "tier": "standard", "quantization": "not reported", "context": 128000, "input": 0.1875, "output": 0.6525, "cacheRead": null, "uptime1d": 99.3 }, { "model": "Llama 4 Maverick", "provider": "Novita", "tier": "standard", "quantization": "fp8", "context": 1048576, "input": 0.27, "output": 0.85, "cacheRead": null, "uptime1d": 97.58 }, { "model": "Llama 4 Maverick", "provider": "Parasail", "tier": "standard", "quantization": "fp8", "context": 524288, "input": 0.35, "output": 1, "cacheRead": 0.17, "uptime1d": 99.87 }, { "model": "Llama 4 Maverick", "provider": "Google Vertex", "tier": "regional", "quantization": "not reported", "context": 524288, "input": 0.35, "output": 1.15, "cacheRead": null, "uptime1d": null }, { "model": "Mistral Large 3 2512", "provider": "Mistral", "tier": "standard", "quantization": "not reported", "context": 262144, "input": 0.5, "output": 1.5, "cacheRead": 0.05, "uptime1d": 99.85 }, { "model": "Mistral Large 3 2512", "provider": "Mistral", "tier": "regional", "quantization": "not reported", "context": 262144, "input": 0.55, "output": 1.65, "cacheRead": 0.055, "uptime1d": 99.84 }, { "model": "Mistral Medium 3.5", "provider": "Mistral", "tier": "standard", "quantization": "not reported", "context": 262144, "input": 1.5, "output": 7.5, "cacheRead": null, "uptime1d": 99.91 }, { "model": "Mistral Medium 3.5", "provider": "Mistral", "tier": "standard", "quantization": "not reported", "context": 262144, "input": 1.5, "output": 7.5, "cacheRead": null, "uptime1d": 99.93 }, { "model": "Mistral Medium 3.5", "provider": "Mistral", "tier": "regional", "quantization": "not reported", "context": 262144, "input": 1.65, "output": 8.25, "cacheRead": null, "uptime1d": 100 }, { "model": "Qwen3.8 Flash", "provider": "Alibaba", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 0.15, "output": 0.47, "cacheRead": 0.016, "uptime1d": 98.7 }, { "model": "Qwen3.8 Max (0902)", "provider": "Alibaba", "tier": "standard", "quantization": "not reported", "context": 1000000, "input": 2, "output": 6, "cacheRead": 0.25, "uptime1d": 99.8 } ] } ], "related": [ "routing-overhead", "cost-thought-experiments" ] }, "sources": [ { "id": "price-anthropic", "title": "Anthropic list prices (Claude models)", "kind": "price-list", "date": "2026-09-21", "url": "https://platform.claude.com/docs/en/about-claude/pricing", "note": "Prices as listed by the vendor on 2026-09-21 and recorded in the product price table. Cache reads at the listed rate, one-hour cache writes at twice the input price." }, { "id": "price-google", "title": "Google Gemini list prices", "kind": "price-list", "date": "2026-09-21", "url": "https://ai.google.dev/pricing", "note": "Gemini 3.x Flash prices as listed by the vendor on 2026-09-21. The vendor announced a doubling from 2027-01-01." }, { "id": "price-openai", "title": "OpenAI list prices", "kind": "price-list", "date": "2026-10-03", "url": "https://developers.openai.com/api/docs/pricing", "note": "Token prices as listed by the vendor on 2026-10-03." }, { "id": "openrouter-api-snapshot", "title": "OpenRouter public API: models and provider endpoints (snapshot)", "kind": "price-list", "date": "2026-10-06", "url": "https://openrouter.ai/docs/api-reference/list-endpoints-for-a-model", "note": "Prices, context, quantization and uptime per provider endpoint as reported by OpenRouter’s public, keyless API on 2026-10-06. Third-party-reported, not measured by Agent. Latency and throughput were not returned.", "data": [ "/benchmarks/raw/provider-index/index.json" ] }, { "id": "openrouter-fees", "title": "OpenRouter pricing and fees", "kind": "price-list", "date": "2026-10-06", "url": "https://openrouter.ai/pricing", "note": "OpenRouter states that inference is billed at the provider list price and that its fee is charged when credits are bought (5.5% on Standard by card, $0.80 minimum; 8% on Business; 5% by crypto). Page fetched 2026-10-06.", "data": [ "/benchmarks/raw/provider-index/index.json" ] } ] }