2026-10-22-provider-pricing

provider,model,processing_tier,modality,context_condition,currency,unit,input_price,output_price,cached_read_price,cache_write_short_price,cache_write_long_price,explicit_cache_storage_per_million_token_hour,region_basis,valid_through,accessed,source_ids,source_url,notes
OpenAI,gpt-6.1-sol,Standard,text,input <=272000,USD,per 1000000 model tokens,2,10,0.1,2.5,5,,default non-regional list,,2026-10-22,C01;C04;C13,https://developers.openai.com/api/docs/pricing,Cache write short=30m; long=12h; reasoning included in output total
OpenAI,gpt-6.1-sol,Batch,text,input <=272000,USD,per 1000000 model tokens,1,5,0.05,1.25,2.5,,default non-regional list,,2026-10-22,C01;C05;C13,https://developers.openai.com/api/docs/pricing,Asynchronous; 24-hour window; cache write short=30m; long=12h
OpenAI,gpt-6-luna,Standard,text,input <=272000,USD,per 1000000 model tokens,0.1,0.5,0.01,0.125,0.25,,default non-regional list,,2026-10-22,C01;C04;C13,https://developers.openai.com/api/docs/pricing,Cache write short=30m; long=12h
Anthropic,Claude Sonnet 5.5,Standard,text,standard full context per current model support,USD,per 1000000 model tokens,2,10,0.1,2.5,4,,first-party global default,,2026-10-22,C02,https://platform.claude.com/docs/en/about-claude/pricing,Cache write short=5m; long=1h
Anthropic,Claude Sonnet 5.5,Batch,text,standard full context per current model support,USD,per 1000000 model tokens,1,5,,,,,first-party global default,,2026-10-22,C02,https://platform.claude.com/docs/en/about-claude/pricing,Cache fields left empty: not separately transcribed; empty is not zero
Anthropic,Claude Haiku 5.5,Standard,text,input <=100000,USD,per 1000000 model tokens,0.1,0.5,0.01,0.125,0.2,,first-party global default,,2026-10-22,C02,https://platform.claude.com/docs/en/about-claude/pricing,Cache write short=5m; long=1h
Anthropic,Claude Haiku 5.5,Standard,text,input >100000,USD,per 1000000 model tokens,0.5,2.5,0.05,0.625,1,,first-party global default,,2026-10-22,C02,https://platform.claude.com/docs/en/about-claude/pricing,All categories use higher long-prompt rates; cache write short=5m; long=1h
Google,gemini-4-argon,Standard paid,text,as listed; implicit cache only,USD,per 1000000 model tokens,2,10,0.1,,,,Gemini Developer API paid,2026-11-30,2026-10-22,C03;C07,https://ai.google.dev/gemini-api/docs/pricing,Promotional rate; announced later prices not included; output includes thinking
Google,gemini-4-argon,Batch paid,text,as listed; implicit cache only,USD,per 1000000 model tokens,1,5,0.05,,,,Gemini Developer API paid,2026-11-30,2026-10-22,C03;C07,https://ai.google.dev/gemini-api/docs/pricing,Promotional rate; explicit cache unsupported
Google,gemini-3.5-flash-lite,Standard paid,text/image/video/audio input; text output,as listed,USD,per 1000000 model tokens,0.3,2.5,0.03,,,1,Gemini Developer API paid,,2026-10-22,C03,https://ai.google.dev/gemini-api/docs/pricing,Explicit storage charge if used; output includes thinking
Google,gemini-3.5-flash-lite,Batch paid,text/image/video/audio input; text output,as listed,USD,per 1000000 model tokens,0.15,1.25,0.02,,,1,Gemini Developer API paid,,2026-10-22,C03,https://ai.google.dev/gemini-api/docs/pricing,Cache price transcribed as displayed; do not derive by halving the Standard read rate