Updated daily · prices checked October 28, 2026
Today's AI model prices
What each company charges per million tokens when you pay per use, grouped by the kind of task each model is good for. Rates here change at midnight Pacific on the day a company's new price takes effect.
Announced price changes: Gemini 4 Argon goes from $2.00 / $10.00 to $4.00 / $20.00 on December 1, 2026; Gemini 3.8 Flash goes from $0.75 / $3.75 to $1.50 / $7.50 on January 1, 2027.
| Task level | Anthropic | OpenAI | |
|---|---|---|---|
| Light Short copy, rewrites, tagging, extracting fields, quick questions | Claude Haiku 5.5 $0.10 in / $0.50 out | GPT-6 Luna $0.10 in / $0.50 out | Gemini 3.1 Flash-Lite $0.25 in / $1.50 out |
| Everyday Emails, blog posts, long summaries, code edits, support replies | Claude Sonnet 5.5 $2.00 in / $10.00 out | GPT-6.1 Terra $1.00 in / $5.00 out | Gemini 3.8 Flash $0.75 in / $3.75 out |
| Hard reasoning Multi-step analysis, complex code, strategy, math | Claude Opus 5.5 $4.00 in / $20.00 out | GPT-6.1 Sol $2.00 in / $10.00 out | Gemini 4 Argon $2.00 in / $10.00 out |
| Frontier Long-running agent work and the hardest problems; rarely needed | Claude Fable 5.1 $10.00 in / $50.00 out | GPT-6 Astra $10.00 in / $50.00 out | None above the tier before |
Prices are list prices per million tokens from each company's pricing page, before discounts such as caching or batch. OpenAI doesn't say what each model is for, so OpenAI models are placed by price until our own tests confirm it. Chat subscriptions (like a $20/month plan) are billed differently; there, using a lighter model means more messages before you hit your limit.