Last updated: 31 October 2026
AI APIs bill per token, quoted per million tokens, with separate input and output prices. On 31 October 2026, standard-tier text prices across OpenAI, Anthropic and Google range from $0.10 input and $0.50 output to $10 input and $50 output per million tokens.
How much does AI inference cost per token?
APIs bill per token, quoted per 1 million tokens, with separate input and output prices; output costs more. Prices change often, so treat the figures below as a snapshot as of 31 October 2026 and check the vendor page before budgeting.
| Model | Input per 1M tokens | Output per 1M tokens |
|---|---|---|
| gpt-6-astra | $10 | $50 |
| gpt-6.1-sol | $2 | $10 |
| gpt-6.1-terra | $1 | $5 |
| gpt-6-luna | $0.10 | $0.50 |
| gpt-5-nano | $0.05 | $0.40 |
| Claude Fable 5.1 | $10 | $50 |
| Claude Opus 5.5 | $4 | $20 |
| Claude Sonnet 5.5 | $2 | $10 |
| Claude Haiku 5.5 (up to 100,000 tokens) | $0.10 | $0.50 |
| Claude Haiku 5.5 (above 100,000 tokens) | $0.50 | $2.50 |
| Gemini 4 Argon (until 30 November 2026) | $2 | $10 |
| Gemini 4 Argon (after 30 November 2026) | $4 | $20 |
| Gemini 3.8 Flash (through 31 December 2026) | $0.75 | $3.75 |
| Gemini 3.8 Flash (after 31 December 2026) | $1.50 | $7.50 |
| Gemini 3.5 Flash-Lite | $0.30 | $2.50 |
| Gemini 2.5 Flash-Lite | $0.10 | $0.40 |
What do OpenAI models cost?
As of 31 October 2026, OpenAI’s listed standard-tier text prices include gpt-6-astra at $10 input and $50 output per million tokens, with cached input at $1. gpt-6.1-sol is $2 input and $10 output, with cached input at $0.10. gpt-6.1-terra is $1 input and $5 output. gpt-6-luna is $0.10 input and $0.50 output. gpt-5-nano is $0.05 input and $0.40 output, an older small model.
What do Anthropic Claude models cost?
As of 31 October 2026, Anthropic’s listed prices include Claude Fable 5.1 at $10 input and $50 output, with cache hits at $0.25. Claude Opus 5.5 is $4 input and $20 output. Claude Sonnet 5.5 is $2 input and $10 output. Claude Haiku 5.5 is $0.10 input and $0.50 output for prompts up to 100,000 tokens, rising to $0.50 input and $2.50 output above 100,000 tokens. Prompt caching writes cost more than base input, with a 5-minute write at $12.50 on Fable 5.1, while cache hits cost far less.
What do Google Gemini models cost?
As of 31 October 2026, Google’s listed prices include Gemini 4 Argon at $2 input and $10 output until 30 November 2026, then $4 input and $20 output after. Gemini 3.8 Flash is $0.75 input and $3.75 output through 31 December 2026, then $1.50 input and $7.50 output after. Gemini 3.5 Flash-Lite is $0.30 input and $2.50 output. Gemini 2.5 Flash-Lite is $0.10 input for text and $0.40 output.
How do you estimate your own cost?
Cost per request equals input tokens divided by 1,000,000, multiplied by the input price, plus output tokens divided by 1,000,000, multiplied by the output price. A worked example: 2,000 input and 500 output tokens at $2 and $10 per million tokens equals $0.004 plus $0.005, or $0.009 per request; 100,000 requests equals $900. Caching, batch discounts and long-context surcharges change the bill, so check the vendor page. Our can help with the rest.
Frequently asked questions
Do output tokens cost more than input tokens?
Yes. Every model in the table above charges more per million output tokens than per million input tokens. Vendors set these prices themselves, so check the pricing page for current detail.
How do I turn tokens into a cost per request?
Use the formula: input tokens divided by 1,000,000 times the input price, plus output tokens divided by 1,000,000 times the output price. For example, 2,000 input and 500 output tokens at $2 and $10 per million tokens cost $0.009 per request.
Are these prices permanent?
No. Prices change often, and some models have stated future price changes, such as Gemini 4 Argon changing after 30 November 2026. Check the vendor page before relying on any figure.
What is cached input?
Cached input is a lower price for reused prompt content, such as $1 on gpt-6-astra or $0.25 cache hits on Claude Fable 5.1. Prompt caching writes can cost more than base input, such as $12.50 for a 5-minute write on Fable 5.1.
Do long prompts change the price?
Yes for at least one model. Claude Haiku 5.5 is $0.10 input and $0.50 output for prompts up to 100,000 tokens, and $0.50 input and $2.50 output above 100,000 tokens. Check the vendor page for other long-context rules.
Sources
Figures and prices were checked on 31 October 2026; check the linked sources before relying on them. This article is general information, not tax, legal or investment advice.
