{"reportId":"llm-state-q3-2026","title":"State of LLM API Pricing and Performance — Q3 2026","asOf":"2026-08-08","datasets":{"pricing":{"id":"all-ai-ask-pricing-registry-2026-08-08","version":"2026-08-08","sourcePath":"backend/src/billing/provider-pricing.registry.ts","sha256":"074ab865dd4fcdf6c261a4213bbb245a1aed6d67540dcdada7dcdc56b5cd71f8"},"speed":{"id":"all-ai-ask-speed-benchmark-2026-08-08","version":"2026-08-08","sourcePath":"backend/src/data/speed-benchmarks.json","sha256":"a6ade1cad8fa7899fdeba7337309db85dcf64130ad4a6aae6865b8074341dc46"}},"methodology":{"pricing":"Registry prices are converted from $/1K tokens to $/1M tokens. Blended price is (3 × input + output) ÷ 4.","speed":"One fixed prompt, 5 runs per model, measured through the All AI Ask gateway from us-east-1."},"tables":{"pricing":[{"modelId":"llama-3.1-8b","model":"Llama 3.1 8B","provider":"Groq","inputPerM":0.05,"outputPerM":0.08,"blendedPerM":0.05750000000000001,"verifiedAt":"2026-04-06","sourceUrl":"https://console.groq.com/docs/models"},{"modelId":"nova-micro","model":"Amazon Nova Micro","provider":"Amazon","inputPerM":0.034999999999999996,"outputPerM":0.13999999999999999,"blendedPerM":0.06124999999999999,"verifiedAt":"2026-06-14","sourceUrl":"https://aws.amazon.com/bedrock/pricing"},{"modelId":"nova-lite","model":"Amazon Nova Lite","provider":"Amazon","inputPerM":0.060000000000000005,"outputPerM":0.24000000000000002,"blendedPerM":0.10500000000000001,"verifiedAt":"2026-06-14","sourceUrl":"https://aws.amazon.com/bedrock/pricing"},{"modelId":"gpt-oss-20b","model":"GPT-OSS 20B","provider":"Groq","inputPerM":0.075,"outputPerM":0.3,"blendedPerM":0.13124999999999998,"verifiedAt":"2026-04-06","sourceUrl":"https://console.groq.com/docs/models"},{"modelId":"gpt-5-nano","model":"GPT-5 Nano","provider":"OpenAI","inputPerM":0.05,"outputPerM":0.4,"blendedPerM":0.1375,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"ministral-8b","model":"Ministral 8B","provider":"Mistral","inputPerM":0.15,"outputPerM":0.15,"blendedPerM":0.15,"verifiedAt":"2026-08-14","sourceUrl":"https://docs.mistral.ai/models/model-cards/ministral-3-8b-25-12"},{"modelId":"gemini-2.5-flash-lite","model":"Gemini 2.5 Flash Lite","provider":"Google","inputPerM":0.1,"outputPerM":0.4,"blendedPerM":0.17500000000000002,"verifiedAt":"2026-04-06","sourceUrl":"https://ai.google.dev/gemini-api/docs/pricing"},{"modelId":"gpt-4o-mini","model":"GPT-4o Mini","provider":"OpenAI","inputPerM":0.15,"outputPerM":0.6,"blendedPerM":0.26249999999999996,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"grok-3-mini","model":"Grok-3 Mini","provider":"xAI","inputPerM":0.15,"outputPerM":0.6,"blendedPerM":0.26249999999999996,"verifiedAt":"2026-04-06","sourceUrl":"https://docs.x.ai/docs/overview#pricing"},{"modelId":"gpt-oss-120b","model":"GPT-OSS 120B","provider":"Groq","inputPerM":0.15,"outputPerM":0.6,"blendedPerM":0.26249999999999996,"verifiedAt":"2026-04-06","sourceUrl":"https://console.groq.com/docs/models"},{"modelId":"mistral-small","model":"Mistral Small 3.1","provider":"Mistral","inputPerM":0.15,"outputPerM":0.6,"blendedPerM":0.26249999999999996,"verifiedAt":"2026-08-14","sourceUrl":"https://docs.mistral.ai/models/model-cards/mistral-small-4-0-26-03"},{"modelId":"llama-4-maverick","model":"Llama 4 Maverick","provider":"Groq","inputPerM":0.2,"outputPerM":0.6,"blendedPerM":0.30000000000000004,"verifiedAt":"2026-07-10","sourceUrl":"https://groq.com/pricing"},{"modelId":"codestral","model":"Codestral","provider":"Mistral","inputPerM":0.3,"outputPerM":0.9,"blendedPerM":0.44999999999999996,"verifiedAt":"2026-06-14","sourceUrl":"https://mistral.ai/pricing"},{"modelId":"cerebras-gpt-oss-120b","model":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","inputPerM":0.35,"outputPerM":0.75,"blendedPerM":0.44999999999999996,"verifiedAt":"2026-06-14","sourceUrl":"https://www.cerebras.ai/pricing"},{"modelId":"gpt-5.4-nano","model":"GPT-5.4 Nano","provider":"OpenAI","inputPerM":0.2,"outputPerM":1.25,"blendedPerM":0.4625,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"gemini-3.1-flash-lite","model":"Gemini 3.1 Flash Lite","provider":"Google","inputPerM":0.25,"outputPerM":1.5,"blendedPerM":0.5625,"verifiedAt":"2026-04-06","sourceUrl":"https://ai.google.dev/gemini-api/docs/pricing"},{"modelId":"llama-3.3-70b","model":"Llama 3.3 70B","provider":"Groq","inputPerM":0.5900000000000001,"outputPerM":0.79,"blendedPerM":0.6400000000000001,"verifiedAt":"2026-04-06","sourceUrl":"https://console.groq.com/docs/models"},{"modelId":"deepseek-v4-flash","model":"DeepSeek V4 Flash","provider":"DeepSeek","inputPerM":0.44,"outputPerM":1.32,"blendedPerM":0.66,"verifiedAt":"2026-08-14","sourceUrl":"https://api-docs.deepseek.com/quick_start/pricing"},{"modelId":"gpt-5-mini","model":"GPT-5 Mini","provider":"OpenAI","inputPerM":0.25,"outputPerM":2,"blendedPerM":0.6875,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"mistral-large","model":"Mistral Large 3","provider":"Mistral","inputPerM":0.5,"outputPerM":1.5,"blendedPerM":0.75,"verifiedAt":"2026-08-14","sourceUrl":"https://docs.mistral.ai/models/model-cards/mistral-large-3-25-12"},{"modelId":"gemini-3.5-flash-lite","model":"Gemini 3.5 Flash Lite","provider":"Google","inputPerM":0.3,"outputPerM":2.5,"blendedPerM":0.85,"verifiedAt":"2026-08-14","sourceUrl":"https://ai.google.dev/gemini-api/docs/latest-model"},{"modelId":"gemini-2.5-flash","model":"Gemini 2.5 Flash","provider":"Google","inputPerM":0.3,"outputPerM":2.5,"blendedPerM":0.85,"verifiedAt":"2026-04-06","sourceUrl":"https://ai.google.dev/gemini-api/docs/pricing"},{"modelId":"glm-5.1","model":"GLM-5.1","provider":"Z.ai","inputPerM":0.6,"outputPerM":2.2,"blendedPerM":1,"verifiedAt":"2026-06-19","sourceUrl":"https://z.ai/model-api"},{"modelId":"qwen3.7-plus","model":"Qwen 3.7 Plus","provider":"Qwen","inputPerM":0.8,"outputPerM":2,"blendedPerM":1.1,"verifiedAt":"2026-07-23","sourceUrl":"https://www.alibabacloud.com/help/en/model-studio/pricing"},{"modelId":"qwen3.8-30b","model":"Qwen 3.8 30B","provider":"Groq","inputPerM":0.6,"outputPerM":3,"blendedPerM":1.2,"verifiedAt":"2026-07-10","sourceUrl":"https://groq.com/pricing"},{"modelId":"qwen3.6-27b","model":"Qwen 3.6 27B","provider":"Groq","inputPerM":0.6,"outputPerM":3,"blendedPerM":1.2,"verifiedAt":"2026-06-19","sourceUrl":"https://groq.com/pricing"},{"modelId":"nova-pro","model":"Amazon Nova Pro","provider":"Amazon","inputPerM":0.8,"outputPerM":3.2,"blendedPerM":1.4000000000000001,"verifiedAt":"2026-06-14","sourceUrl":"https://aws.amazon.com/bedrock/pricing"},{"modelId":"gemini-3.7-flash","model":"Gemini 3.7 Flash","provider":"Google","inputPerM":0.75,"outputPerM":3.75,"blendedPerM":1.5,"verifiedAt":"2026-08-14","sourceUrl":"https://ai.google.dev/gemini-api/docs/pricing"},{"modelId":"grok-4.3","model":"Grok 4.3","provider":"xAI","inputPerM":1.25,"outputPerM":2.5,"blendedPerM":1.5625,"verifiedAt":"2026-05-19","sourceUrl":"https://docs.x.ai/docs/overview#pricing"},{"modelId":"gpt-5.4-mini","model":"GPT-5.4 Mini","provider":"OpenAI","inputPerM":0.75,"outputPerM":4.5,"blendedPerM":1.6875,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"gemini-3.1-flash","model":"Gemini 3.1 Flash","provider":"Google","inputPerM":0.75,"outputPerM":4.5,"blendedPerM":1.6875,"verifiedAt":"2026-04-06","sourceUrl":"https://ai.google.dev/gemini-api/docs/pricing"},{"modelId":"o3-mini","model":"o3-Mini","provider":"OpenAI","inputPerM":1.1,"outputPerM":4.4,"blendedPerM":1.9250000000000003,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"deepseek-v4-pro","model":"DeepSeek V4 Pro","provider":"DeepSeek","inputPerM":1.32,"outputPerM":3.96,"blendedPerM":1.98,"verifiedAt":"2026-08-14","sourceUrl":"https://api-docs.deepseek.com/quick_start/pricing"},{"modelId":"claude-haiku-4-5","model":"Claude Haiku 4.5","provider":"Anthropic","inputPerM":1,"outputPerM":5,"blendedPerM":2,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"glm-5.2","model":"GLM-5.2","provider":"Z.ai","inputPerM":1.4,"outputPerM":4.4,"blendedPerM":2.15,"verifiedAt":"2026-06-19","sourceUrl":"https://z.ai/model-api"},{"modelId":"gpt-5.6-luna","model":"GPT-5.6 Luna","provider":"OpenAI","inputPerM":1,"outputPerM":6,"blendedPerM":2.25,"verifiedAt":"2026-08-14","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"cerebras-glm-4.7","model":"GLM 4.7 (Cerebras)","provider":"Cerebras","inputPerM":2.25,"outputPerM":2.75,"blendedPerM":2.375,"verifiedAt":"2026-06-14","sourceUrl":"https://www.cerebras.ai/pricing"},{"modelId":"grok-3","model":"Grok-3","provider":"xAI","inputPerM":2,"outputPerM":4,"blendedPerM":2.5,"verifiedAt":"2026-04-06","sourceUrl":"https://docs.x.ai/docs/overview#pricing"},{"modelId":"qwen3.8-max","model":"Qwen 3.8 Max","provider":"Qwen","inputPerM":1.6,"outputPerM":6.4,"blendedPerM":2.8000000000000003,"verifiedAt":"2026-07-10","sourceUrl":"https://www.alibabacloud.com/help/en/model-studio/pricing"},{"modelId":"qwen3.7-max","model":"Qwen 3.7 Max","provider":"Qwen","inputPerM":1.6,"outputPerM":6.4,"blendedPerM":2.8000000000000003,"verifiedAt":"2026-07-23","sourceUrl":"https://www.alibabacloud.com/help/en/model-studio/pricing"},{"modelId":"grok-4.20-0309-reasoning","model":"Grok-4.20 Reasoning","provider":"xAI","inputPerM":2,"outputPerM":6,"blendedPerM":3,"verifiedAt":"2026-04-06","sourceUrl":"https://docs.x.ai/docs/overview#pricing"},{"modelId":"grok-4.20-0309-non-reasoning","model":"Grok-4.20","provider":"xAI","inputPerM":2,"outputPerM":6,"blendedPerM":3,"verifiedAt":"2026-04-06","sourceUrl":"https://docs.x.ai/docs/overview#pricing"},{"modelId":"grok-4.6","model":"Grok 4.6","provider":"xAI","inputPerM":2,"outputPerM":6,"blendedPerM":3,"verifiedAt":"2026-08-14","sourceUrl":"https://docs.x.ai/docs/models"},{"modelId":"grok-4.5","model":"Grok 4.5","provider":"xAI","inputPerM":2,"outputPerM":6,"blendedPerM":3,"verifiedAt":"2026-08-14","sourceUrl":"https://docs.x.ai/developers/models/grok-4.5"},{"modelId":"gemini-3.6-flash","model":"Gemini 3.6 Flash","provider":"Google","inputPerM":1.5,"outputPerM":7.5,"blendedPerM":3,"verifiedAt":"2026-08-14","sourceUrl":"https://ai.google.dev/gemini-api/docs/latest-model"},{"modelId":"mistral-medium","model":"Mistral Medium 3","provider":"Mistral","inputPerM":1.5,"outputPerM":7.5,"blendedPerM":3,"verifiedAt":"2026-08-14","sourceUrl":"https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04"},{"modelId":"gemini-3.5-flash","model":"Gemini 3.5 Flash","provider":"Google","inputPerM":1.5,"outputPerM":9,"blendedPerM":3.375,"verifiedAt":"2026-08-14","sourceUrl":"https://ai.google.dev/gemini-api/docs/pricing"},{"modelId":"gpt-5","model":"GPT-5","provider":"OpenAI","inputPerM":1.25,"outputPerM":10,"blendedPerM":3.4375,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"gpt-4.1","model":"GPT-4.1","provider":"OpenAI","inputPerM":2,"outputPerM":8,"blendedPerM":3.5,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"claude-sonnet-5","model":"Claude Sonnet 5","provider":"Anthropic","inputPerM":2,"outputPerM":10,"blendedPerM":4,"verifiedAt":"2026-08-14","sourceUrl":"https://www.anthropic.com/news/claude-sonnet-5"},{"modelId":"gpt-4o","model":"GPT-4o","provider":"OpenAI","inputPerM":2.5,"outputPerM":10,"blendedPerM":4.375,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"gemini-3.1-pro","model":"Gemini 3.1 Pro","provider":"Google","inputPerM":2,"outputPerM":12,"blendedPerM":4.5,"verifiedAt":"2026-04-06","sourceUrl":"https://ai.google.dev/gemini-api/docs/pricing"},{"modelId":"gpt-5.6-terra","model":"GPT-5.6 Terra","provider":"OpenAI","inputPerM":2.5,"outputPerM":15,"blendedPerM":5.625,"verifiedAt":"2026-08-14","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"gpt-5.4","model":"GPT-5.4","provider":"OpenAI","inputPerM":2.5,"outputPerM":15,"blendedPerM":5.625,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"claude-sonnet-4-6","model":"Claude Sonnet 4.6","provider":"Anthropic","inputPerM":3,"outputPerM":15,"blendedPerM":6,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-sonnet-4-5","model":"Claude Sonnet 4.5","provider":"Anthropic","inputPerM":3,"outputPerM":15,"blendedPerM":6,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-sonnet-4","model":"Claude Sonnet 4","provider":"Anthropic","inputPerM":3,"outputPerM":15,"blendedPerM":6,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-opus-4-8","model":"Claude Opus 4.8","provider":"Anthropic","inputPerM":5,"outputPerM":25,"blendedPerM":10,"verifiedAt":"2026-06-07","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-opus-4-7","model":"Claude Opus 4.7","provider":"Anthropic","inputPerM":5,"outputPerM":25,"blendedPerM":10,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-opus-4-6","model":"Claude Opus 4.6","provider":"Anthropic","inputPerM":5,"outputPerM":25,"blendedPerM":10,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-opus-4-5","model":"Claude Opus 4.5","provider":"Anthropic","inputPerM":5,"outputPerM":25,"blendedPerM":10,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"gpt-5.6-sol","model":"GPT-5.6 Sol","provider":"OpenAI","inputPerM":5,"outputPerM":30,"blendedPerM":11.25,"verifiedAt":"2026-08-14","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"gpt-4-turbo","model":"GPT-4 Turbo","provider":"OpenAI","inputPerM":10,"outputPerM":30,"blendedPerM":15,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"},{"modelId":"claude-fable-5","model":"Claude Fable 5","provider":"Anthropic","inputPerM":10,"outputPerM":50,"blendedPerM":20,"verifiedAt":"2026-08-14","sourceUrl":"https://www.anthropic.com/claude/fable"},{"modelId":"claude-opus-5","model":"Claude Opus 5","provider":"Anthropic","inputPerM":15,"outputPerM":75,"blendedPerM":30,"verifiedAt":"2026-08-09","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-opus-4-1","model":"Claude Opus 4.1","provider":"Anthropic","inputPerM":15,"outputPerM":75,"blendedPerM":30,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"claude-opus-4","model":"Claude Opus 4","provider":"Anthropic","inputPerM":15,"outputPerM":75,"blendedPerM":30,"verifiedAt":"2026-04-06","sourceUrl":"https://www.anthropic.com/pricing"},{"modelId":"gpt-5.4-pro","model":"GPT-5.4 Pro","provider":"OpenAI","inputPerM":30,"outputPerM":180,"blendedPerM":67.5,"verifiedAt":"2026-04-06","sourceUrl":"https://openai.com/api/pricing"}],"speed":[{"modelId":"cerebras-gpt-oss-120b","model":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","rank":1,"ttftMs":90,"tokensPerSecond":2450,"p95TokensPerSecond":2082.5,"samples":5,"blendedPerM":0.44999999999999996,"costPerMTokPerSec":0.0002},{"modelId":"cerebras-glm-4.7","model":"GLM 4.7 (Cerebras)","provider":"Cerebras","rank":2,"ttftMs":110,"tokensPerSecond":1980,"p95TokensPerSecond":1702.8,"samples":5,"blendedPerM":2.375,"costPerMTokPerSec":0.0012},{"modelId":"gpt-oss-20b","model":"GPT-OSS 20B","provider":"Groq","rank":3,"ttftMs":140,"tokensPerSecond":1120,"p95TokensPerSecond":974.4,"samples":5,"blendedPerM":0.13124999999999998,"costPerMTokPerSec":0.0001},{"modelId":"gpt-oss-120b","model":"GPT-OSS 120B","provider":"Groq","rank":4,"ttftMs":160,"tokensPerSecond":780,"p95TokensPerSecond":639.6,"samples":5,"blendedPerM":0.26249999999999996,"costPerMTokPerSec":0.0003},{"modelId":"qwen3.8-30b","model":"Qwen 3.8 30B","provider":"Groq","rank":5,"ttftMs":150,"tokensPerSecond":690,"p95TokensPerSecond":634.8,"samples":5,"blendedPerM":1.2,"costPerMTokPerSec":0.0017},{"modelId":"nova-micro","model":"Amazon Nova Micro","provider":"Amazon","rank":6,"ttftMs":220,"tokensPerSecond":168,"p95TokensPerSecond":142.8,"samples":5,"blendedPerM":0.06124999999999999,"costPerMTokPerSec":0.0004},{"modelId":"gemini-3.5-flash-lite","model":"Gemini 3.5 Flash Lite","provider":"Google","rank":7,"ttftMs":240,"tokensPerSecond":162,"p95TokensPerSecond":129.6,"samples":5,"blendedPerM":0.85,"costPerMTokPerSec":0.0052},{"modelId":"ministral-8b","model":"Ministral 8B","provider":"Mistral","rank":8,"ttftMs":210,"tokensPerSecond":158,"p95TokensPerSecond":131.14,"samples":5,"blendedPerM":0.15,"costPerMTokPerSec":0.0009},{"modelId":"claude-haiku-4-5","model":"Claude Haiku 4.5","provider":"Anthropic","rank":9,"ttftMs":260,"tokensPerSecond":148,"p95TokensPerSecond":118.4,"samples":5,"blendedPerM":2,"costPerMTokPerSec":0.0135},{"modelId":"deepseek-v4-flash","model":"DeepSeek V4 Flash","provider":"DeepSeek","rank":10,"ttftMs":280,"tokensPerSecond":132,"p95TokensPerSecond":118.8,"samples":5,"blendedPerM":0.66,"costPerMTokPerSec":0.005},{"modelId":"gpt-5.6-luna","model":"GPT-5.6 Luna","provider":"OpenAI","rank":11,"ttftMs":300,"tokensPerSecond":126,"p95TokensPerSecond":108.36,"samples":5,"blendedPerM":2.25,"costPerMTokPerSec":0.0179},{"modelId":"mistral-small","model":"Mistral Small 3.1","provider":"Mistral","rank":12,"ttftMs":260,"tokensPerSecond":121,"p95TokensPerSecond":110.11,"samples":5,"blendedPerM":0.26249999999999996,"costPerMTokPerSec":0.0022},{"modelId":"codestral","model":"Codestral","provider":"Mistral","rank":13,"ttftMs":270,"tokensPerSecond":118,"p95TokensPerSecond":99.12,"samples":5,"blendedPerM":0.44999999999999996,"costPerMTokPerSec":0.0038},{"modelId":"gemini-3.6-flash","model":"Gemini 3.6 Flash","provider":"Google","rank":14,"ttftMs":310,"tokensPerSecond":114,"p95TokensPerSecond":93.48,"samples":5,"blendedPerM":3,"costPerMTokPerSec":0.0263},{"modelId":"nova-lite","model":"Amazon Nova Lite","provider":"Amazon","rank":15,"ttftMs":250,"tokensPerSecond":108,"p95TokensPerSecond":95.04,"samples":5,"blendedPerM":0.10500000000000001,"costPerMTokPerSec":0.001},{"modelId":"grok-4.20-0309-non-reasoning","model":"Grok-4.20","provider":"xAI","rank":16,"ttftMs":290,"tokensPerSecond":104,"p95TokensPerSecond":91.52,"samples":5,"blendedPerM":3,"costPerMTokPerSec":0.0288},{"modelId":"grok-4.3","model":"Grok 4.3","provider":"xAI","rank":17,"ttftMs":320,"tokensPerSecond":98,"p95TokensPerSecond":82.32,"samples":5,"blendedPerM":1.5625,"costPerMTokPerSec":0.0159},{"modelId":"mistral-medium","model":"Mistral Medium 3","provider":"Mistral","rank":18,"ttftMs":320,"tokensPerSecond":92,"p95TokensPerSecond":73.6,"samples":5,"blendedPerM":3,"costPerMTokPerSec":0.0326},{"modelId":"qwen3.7-plus","model":"Qwen 3.7 Plus","provider":"Qwen","rank":19,"ttftMs":340,"tokensPerSecond":84,"p95TokensPerSecond":75.6,"samples":5,"blendedPerM":1.1,"costPerMTokPerSec":0.0131},{"modelId":"gpt-5.6-terra","model":"GPT-5.6 Terra","provider":"OpenAI","rank":20,"ttftMs":380,"tokensPerSecond":78,"p95TokensPerSecond":63.96,"samples":5,"blendedPerM":5.625,"costPerMTokPerSec":0.0721},{"modelId":"claude-sonnet-4-6","model":"Claude Sonnet 4.6","provider":"Anthropic","rank":21,"ttftMs":360,"tokensPerSecond":76,"p95TokensPerSecond":69.92,"samples":5,"blendedPerM":6,"costPerMTokPerSec":0.0789},{"modelId":"deepseek-v4-pro","model":"DeepSeek V4 Pro","provider":"DeepSeek","rank":22,"ttftMs":480,"tokensPerSecond":68,"p95TokensPerSecond":62.56,"samples":5,"blendedPerM":1.98,"costPerMTokPerSec":0.0291},{"modelId":"nova-pro","model":"Amazon Nova Pro","provider":"Amazon","rank":23,"ttftMs":350,"tokensPerSecond":64,"p95TokensPerSecond":55.04,"samples":5,"blendedPerM":1.4000000000000001,"costPerMTokPerSec":0.0219},{"modelId":"mistral-large","model":"Mistral Large 3","provider":"Mistral","rank":24,"ttftMs":400,"tokensPerSecond":61,"p95TokensPerSecond":50.02,"samples":5,"blendedPerM":0.75,"costPerMTokPerSec":0.0123},{"modelId":"claude-opus-4-8","model":"Claude Opus 4.8","provider":"Anthropic","rank":25,"ttftMs":470,"tokensPerSecond":58,"p95TokensPerSecond":53.36,"samples":5,"blendedPerM":10,"costPerMTokPerSec":0.1724},{"modelId":"gemini-3.1-pro","model":"Gemini 3.1 Pro","provider":"Google","rank":26,"ttftMs":420,"tokensPerSecond":55,"p95TokensPerSecond":50.6,"samples":5,"blendedPerM":4.5,"costPerMTokPerSec":0.0818},{"modelId":"grok-4.20-0309-reasoning","model":"Grok-4.20 Reasoning","provider":"xAI","rank":27,"ttftMs":540,"tokensPerSecond":52,"p95TokensPerSecond":46.8,"samples":5,"blendedPerM":3,"costPerMTokPerSec":0.0577},{"modelId":"qwen3.7-max","model":"Qwen 3.7 Max","provider":"Qwen","rank":28,"ttftMs":460,"tokensPerSecond":49,"p95TokensPerSecond":45.08,"samples":5,"blendedPerM":2.8000000000000003,"costPerMTokPerSec":0.0571},{"modelId":"qwen3.8-max","model":"Qwen 3.8 Max","provider":"Qwen","rank":29,"ttftMs":470,"tokensPerSecond":47,"p95TokensPerSecond":37.6,"samples":5,"blendedPerM":2.8000000000000003,"costPerMTokPerSec":0.0596},{"modelId":"gpt-5.6-sol","model":"GPT-5.6 Sol","provider":"OpenAI","rank":30,"ttftMs":560,"tokensPerSecond":44,"p95TokensPerSecond":35.64,"samples":5,"blendedPerM":11.25,"costPerMTokPerSec":0.2557},{"modelId":"claude-fable-5","model":"Claude Fable 5","provider":"Anthropic","rank":31,"ttftMs":610,"tokensPerSecond":41,"p95TokensPerSecond":35.67,"samples":5,"blendedPerM":20,"costPerMTokPerSec":0.4878}]}}