AI Compare · API costs
What AI models cost through their APIs
Enter your workload and see what a month of use would cost on each model, at the list prices each company publishes.
Prices for United States, before sales tax Data release 2026.10.16-3, Oct 16, 2026 Rules 2.1.0
If you build an app on an AI model, you pay per token: a share of a word, counted in what you send (input) and what the model writes back (output, including its hidden reasoning). This page uses each company's own list prices. Enter your workload to see a month's cost on every model.
Your workload
The same text becomes a different number of tokens with each company: Anthropic says its newer models count about 30% more tokens than its older ones. Compare with token counts from your own logs where you can. A lower price says nothing about the quality of the answers.
List prices
USD per 1M tokens. Standard pay-as-you-go list prices from each company's own pricing page, global endpoints. Not included: batch, flex and priority tiers, regional uplifts, tool fees such as web search, free tiers and volume discounts.
| Model | Input | Cached input | Cache write | Output | Notes |
|---|---|---|---|---|---|
| GPT-6 Astragpt-6-astra · OpenAI | $10.00 | $1.00 | $12.50 | $50.00 | Over 272,000 input tokens: input $20.00, output $75.00. Source |
| GPT-6.1 Solgpt-6.1-sol · OpenAI | $2.00 | $0.10 | $2.50 | $10.00 | Over 272,000 input tokens: input $4.00, output $15.00. Source |
| GPT-6 Lunagpt-6-luna · OpenAI | $0.10 | $0.01 | $0.125 | $0.50 | Over 272,000 input tokens: input $0.20, output $0.75. Source |
| GPT-6 Solgpt-6-sol · OpenAI | $2.00 | $0.20 | $2.50 | $10.00 | Long-context prices (over 272K input tokens) are not listed for this model on the pricing page. Source |
| GPT-5.6 Solgpt-5.6-sol · OpenAI | $4.00 | $0.40 | $5.00 | $20.00 | Over 272,000 input tokens: input $8.00, output $30.00. The pricing page says GPT-5.6 Sol's promotional pricing is available at least through November 21, 2026. Source |
| GPT-5.6 Terragpt-5.6-terra · OpenAI | $2.00 | $0.20 | $2.50 | $12.00 | Long-context prices (over 272K input tokens) are not listed for this model on the pricing page. Source |
| GPT-5.6 Lunagpt-5.6-luna · OpenAI | $0.20 | $0.02 | $0.25 | $1.20 | Long-context prices (over 272K input tokens) are not listed for this model on the pricing page. Source |
| Claude Fable 5.1claude-fable-5-1 · Anthropic | $10 | $0.25 | $12.50 | $50 | Cache writes priced at the 5-minute rate; 1-hour cache writes cost more. Claude 4.7 and later models use a tokenizer that produces about 30% more tokens for the same text. The full 1M-token context window is billed at the same rates. Source |
| Claude Opus 5.5claude-opus-5-5 · Anthropic | $4 | $0.20 | $5 | $20 | Cache writes priced at the 5-minute rate; 1-hour cache writes cost more. Claude 4.7 and later models use a tokenizer that produces about 30% more tokens for the same text. The full 1M-token context window is billed at the same rates. Source |
| Claude Sonnet 5.5claude-sonnet-5-5 · Anthropic | $2 | $0.10 | $2.50 | $10 | Cache writes priced at the 5-minute rate; 1-hour cache writes cost more. Claude 4.7 and later models use a tokenizer that produces about 30% more tokens for the same text. The full 1M-token context window is billed at the same rates. Source |
| Claude Haiku 5.5claude-haiku-5-5 · Anthropic | $0.10 | $0.01 | $0.125 | $0.50 | Over 100,000 input tokens: input $0.50, output $2.50. Cache writes priced at the 5-minute rate; 1-hour cache writes cost more. Claude 4.7 and later models use a tokenizer that produces about 30% more tokens for the same text. Source |
| Gemini 3.1 Pro (preview)gemini-3.1-pro-preview · Google · preview | $2.00 | $0.20 | — | $12.00 | Over 200,000 input tokens: input $4.00, output $18.00. Cached tokens also pay storage: $4.50 per 1M tokens per hour, not included here. Output prices include thinking tokens. Source |
| Gemini 3.8 Flashgemini-3.8-flash · Google | $0.75 | $0.075 | — | $3.75 | From Jan 1, 2027: input $1.50, output $7.50. Cached tokens also pay storage: $0.50 (from January 1, 2027, $1.00) per 1M tokens per hour, not included here. Output prices include thinking tokens. Source |
| Gemini 3.6 Flashgemini-3.6-flash · Google | $0.75 | $0.075 | — | $3.75 | From Jan 1, 2027: input $1.50, output $7.50. Cached tokens also pay storage: $0.50 (from January 1, 2027, $1.00) per 1M tokens per hour, not included here. Output prices include thinking tokens. Source |
| Gemini 3.5 Flash-Litegemini-3.5-flash-lite · Google | $0.30 | $0.03 | — | $2.50 | Cached tokens also pay storage: $1.00 per 1M tokens per hour, not included here. Output prices include thinking tokens. Source |
"—" means the company lists no separate price for it. Checked Oct 13, 2026. Perplexity's API sells other companies' models at their published rates plus tool fees, so it has no separate rows here.