AI API prices, per million tokens
What developers pay to use each provider's models through its API, read on the provider's own pricing page. Standard prices per 1 million tokens, in the currency listed, not converted or rounded. The note under a model says when long prompts, batch use or time of day change the price.
46 models from 10 providers · checked 2026-10-10
New to this? What tokens and these prices mean
A token is a piece of text, often part of a word; a page of English is very roughly 500 to 700 tokens. Input is what you send to the model, output is what it writes back, and cached input is a lower price some providers charge when you send the same start of a prompt again.
The context window is how much text the model can take in at once. If you only chat in an app such as ChatGPT or Claude, you pay for a plan instead; see Compare.
What each AI costs
Each chat app's own plans, next to what the same company charges through its API for the use you pick.
Pick how much you chat
Plans cost the same however much you use them. The API bills by use, so its price follows your choice.
Messages a day
Answers
30 messages a day with short answers is about 660K tokens in and 180K out a month.
ChatGPTOpenAI API$0.16 – $52 a month through the API, 8 models.App plans: Free plan · Go $8.00 / month (United States)
- GPT-6 Luna$0.10 in · $0.50 out per 1M tokensabout $0.16 a month
- GPT-5.6 Luna$0.20 in · $1.20 out per 1M tokensabout $0.35 a month
- GPT-6.1 Sol$2.00 in · $10.00 out per 1M tokensabout $3.12 a month
- GPT-6 Sol$2.00 in · $10.00 out per 1M tokensabout $3.12 a month
- GPT-5.6 Terra$2.00 in · $12.00 out per 1M tokensabout $3.48 a month
- GPT-5.6 Sol$4.00 in · $20.00 out per 1M tokensabout $6.24 a month
- GPT-6 Astra$10.00 in · $50.00 out per 1M tokensabout $16 a month
- GPT-5.5 Pro$30.00 in · $180.00 out per 1M tokensabout $52 a month
ClaudeAnthropic API$0.16 – $16 a month through the API, 4 models.App plans: Free plan · Pro $20.00 / month
- Claude Haiku 5.5$0.10 in · $0.50 out per 1M tokensabout $0.16 a month
- Claude Sonnet 5.5$2.00 in · $10.00 out per 1M tokensabout $3.12 a month
- Claude Opus 5.5$4.00 in · $20.00 out per 1M tokensabout $6.24 a month
- Claude Fable 5.1$10.00 in · $50.00 out per 1M tokensabout $16 a month
DeepSeekDeepSeek API$0.41 – $1.58 a month through the API, 2 models.App plans: Free plan
- DeepSeek-V4.1-Flash$0.30 in · $1.20 out per 1M tokensabout $0.41 a month
- DeepSeek-V4-Pro-0813$1.32 in · $3.96 out per 1M tokensabout $1.58 a month
GeminiGoogle API$0.14 – $3.48 a month through the API, 7 models.App plans: Free plan · Google AI Plus $4.99 / month
- Gemini 2.5 Flash-Lite$0.10 in · $0.40 out per 1M tokensabout $0.14 a month
- Gemini 3.5 Flash-Lite$0.30 in · $2.50 out per 1M tokensabout $0.65 a month
- Gemini 2.5 Flash$0.30 in · $2.50 out per 1M tokensabout $0.65 a month
- Gemini 3.8 Flash$0.75 in · $3.75 out per 1M tokensabout $1.17 a month
- Gemini 3.6 Flash$0.75 in · $3.75 out per 1M tokensabout $1.17 a month
- Gemini 2.5 Pro$1.25 in · $10.00 out per 1M tokensabout $2.63 a month
- Gemini 3.1 Pro Preview$2.00 in · $12.00 out per 1M tokensabout $3.48 a month
GrokxAI API$1.02 – $2.40 a month through the API, 6 models.App plans: Free plan · SuperGrok $30.00 / month
- Grok Build 0.1$1.00 in · $2.00 out per 1M tokensabout $1.02 a month
- Grok 4.3$1.25 in · $2.50 out per 1M tokensabout $1.28 a month
- Grok 4.20$1.25 in · $2.50 out per 1M tokensabout $1.28 a month
- Grok 4.7$2.00 in · $6.00 out per 1M tokensabout $2.40 a month
- Grok 4.6$2.00 in · $6.00 out per 1M tokensabout $2.40 a month
- Grok 4.5$2.00 in · $6.00 out per 1M tokensabout $2.40 a month
KimiMoonshot AI API$1.35 – $4.68 a month through the API, 4 models.App plans: Free plan · Plus $19.00 / month
- Kimi K2.7 Code$0.95 in · $4.00 out per 1M tokensabout $1.35 a month
- Kimi K2.6$0.95 in · $4.00 out per 1M tokensabout $1.35 a month
- Kimi K2.7 Code HighSpeed$1.90 in · $8.00 out per 1M tokensabout $2.69 a month
- Kimi K3$3.00 in · $15.00 out per 1M tokensabout $4.68 a month
Mistral VibeMistral AI API$0.17 – $2.34 a month through the API, 6 models.App plans: Free plan · Pro $14.99 / month
- Ministral 3 14B$0.20 in · $0.20 out per 1M tokensabout $0.17 a month
- Mistral Small 4$0.15 in · $0.60 out per 1M tokensabout $0.21 a month
- Codestral$0.30 in · $0.90 out per 1M tokensabout $0.36 a month
- Mistral Large 3$0.50 in · $1.50 out per 1M tokensabout $0.60 a month
- Mistral Large 4$1.36 in · $4.18 out per 1M tokensabout $1.65 a month
- Mistral Medium 3.5$1.50 in · $7.50 out per 1M tokensabout $2.34 a month
Qwen ChatAlibaba Cloud API$0.04 – $2.40 a month through the API, 4 models.App plans: Free plan
- qwen3.7-flash$0.03 in · $0.13 out per 1M tokensabout $0.04 a month
- qwen3.8-flash$0.15 in · $0.47 out per 1M tokensabout $0.18 a month
- qwen3.7-plus$0.40 in · $1.60 out per 1M tokensabout $0.55 a month
- qwen3.8-max$2.00 in · $6.00 out per 1M tokensabout $2.40 a month
An estimate for text chats, not a quote. Models that “think” before answering, files and images cost more. Each dot is one model; open a row to see them. Prices from the official price pages of 8 companies, checked 2026-10-10.
How we count
- Each message: you write about 50 words; the answer is about 150 words, or 600 for long answers.
- Chats run about 6 messages. A chat API is paid again for the earlier messages each time you send a new one, so on average each message also carries 2.5 earlier questions and answers.
- 100 tokens are about 75 English words, OpenAI's rule of thumb. Other companies count a little differently.
- 30 days a month. Standard prices per 1 million tokens, without caching, batch or off-peak discounts. The price notes on the API price page say when long prompts cost more.
- Not counted: the hidden “thinking” that reasoning models also bill as output, files, images, voice, web search and tax.
- The API needs an account with the company, a key, and an app or code that uses it. An app plan includes the app itself and its own limits.
Every model and price
| Model | Cached input | Context | Source | |||
|---|---|---|---|---|---|---|
| qwen3.7-flashqwen3.7-flashInternational (Singapore) price for prompts up to 32K tokens; longer prompts cost up to $0.200 input, $0.800 output. | Alibaba Cloud (Qwen) | $0.03 | $0.13 | Not listed | 1M | alibabacloud.comchecked 2026-10-10 |
| qwen3.7-plusqwen3.7-plusInternational (Singapore) price for prompts up to 256K tokens; above that $1.2 input, $4.8 output. Listed with a limited-time 20% discount. | Alibaba Cloud (Qwen) | $0.40 | $1.60 | Not listed | 1M | alibabacloud.comchecked 2026-10-10 |
| qwen3.8-flashqwen3.8-flashInternational (Singapore) price. | Alibaba Cloud (Qwen) | $0.15 | $0.47 | Not listed | 1M | alibabacloud.comchecked 2026-10-10 |
| qwen3.8-maxqwen3.8-maxInternational (Singapore) price. | Alibaba Cloud (Qwen) | $2.00 | $6.00 | Not listed | 1M | alibabacloud.comchecked 2026-10-10 |
| Claude Fable 5.1claude-fable-5-1 | Anthropic | $10.00 | $50.00 | $0.25 | 1M | platform.claude.comchecked 2026-10-10 |
| Claude Haiku 5.5claude-haiku-5-5Prompts over 100,000 tokens: $0.50 input, $2.50 output for the whole request. | Anthropic | $0.10 | $0.50 | $0.01 | 1M | platform.claude.comchecked 2026-10-10 |
| Claude Opus 5.5claude-opus-5-5 | Anthropic | $4.00 | $20.00 | $0.20 | 1M | platform.claude.comchecked 2026-10-10 |
| Claude Sonnet 5.5claude-sonnet-5-5 | Anthropic | $2.00 | $10.00 | $0.10 | 1M | platform.claude.comchecked 2026-10-10 |
| DeepSeek-V4-Pro-0813deepseek-v4-proPeak-hour price. Off-peak it is half: $0.66 input, $1.98 output. | DeepSeek | $1.32 | $3.96 | $0.044 | 1M | api-docs.deepseek.comchecked 2026-10-10 |
| DeepSeek-V4.1-Flashdeepseek-flashPeak-hour price. Off-peak it is half: $0.15 input, $0.6 output. | DeepSeek | $0.30 | $1.20 | $0.006 | 1M | api-docs.deepseek.comchecked 2026-10-10 |
| Gemini 2.5 Flashgemini-2.5-flashAudio input costs more: $1.00. | $0.30 | $2.50 | $0.03 | 1.05M | ai.google.devchecked 2026-10-10 | |
| Gemini 2.5 Flash-Litegemini-2.5-flash-liteAudio input costs more: $0.30. | $0.10 | $0.40 | $0.01 | 1.05M | ai.google.devchecked 2026-10-10 | |
| Gemini 2.5 Progemini-2.5-proPrompts over 200K tokens: $2.50 input, $15.00 output. | $1.25 | $10.00 | $0.125 | 1.05M | ai.google.devchecked 2026-10-10 | |
| Gemini 3.1 Pro Previewgemini-3.1-pro-previewPreview model. Prompts over 200K tokens: $4.00 input, $18.00 output. | $2.00 | $12.00 | $0.20 | 1.05M | ai.google.devchecked 2026-10-10 | |
| Gemini 3.5 Flash-Litegemini-3.5-flash-lite | $0.30 | $2.50 | $0.03 | 1.05M | ai.google.devchecked 2026-10-10 | |
| Gemini 3.6 Flashgemini-3.6-flashPrice valid through December 31, 2026. From January 1, 2027: $1.50 input, $7.50 output. | $0.75 | $3.75 | $0.075 | 1.05M | ai.google.devchecked 2026-10-10 | |
| Gemini 3.8 Flashgemini-3.8-flashPrice valid through December 31, 2026. From January 1, 2027: $1.50 input, $7.50 output. | $0.75 | $3.75 | $0.075 | 1.05M | ai.google.devchecked 2026-10-10 | |
| Codestralcodestral-2508Model for code. | Mistral AI | $0.30 | $0.90 | $0.03 | 128K | docs.mistral.aichecked 2026-10-10 |
| Ministral 3 14Bministral-14b-2512 | Mistral AI | $0.20 | $0.20 | $0.02 | 256K | docs.mistral.aichecked 2026-10-10 |
| Mistral Large 3mistral-large-2512 | Mistral AI | $0.50 | $1.50 | $0.05 | 256K | docs.mistral.aichecked 2026-10-10 |
| Mistral Large 4mistral-large-4Public preview. The pricing page currently shows a sale price of $0.68 input, $2.09 output, with no end date. | Mistral AI | $1.36 | $4.18 | $0.14 | 1M | docs.mistral.aichecked 2026-10-10 |
| Mistral Medium 3.5mistral-medium-3-5 | Mistral AI | $1.50 | $7.50 | $0.15 | 256K | docs.mistral.aichecked 2026-10-10 |
| Mistral Small 4mistral-small-2603 | Mistral AI | $0.15 | $0.60 | $0.015 | 256K | docs.mistral.aichecked 2026-10-10 |
| Kimi K2.6kimi-k2.6 | Moonshot AI (Kimi) | $0.95 | $4.00 | $0.16 | 262.1K | platform.kimi.aichecked 2026-10-10 |
| Kimi K2.7 Codekimi-k2.7-code | Moonshot AI (Kimi) | $0.95 | $4.00 | $0.19 | 262.1K | platform.kimi.aichecked 2026-10-10 |
| Kimi K2.7 Code HighSpeedkimi-k2.7-code-highspeed | Moonshot AI (Kimi) | $1.90 | $8.00 | $0.38 | 262.1K | platform.kimi.aichecked 2026-10-10 |
| Kimi K3kimi-k3Writing to the cache costs extra. | Moonshot AI (Kimi) | $3.00 | $15.00 | $0.30 | 1.05M | platform.kimi.aichecked 2026-10-10 |
| GPT-5.5 Progpt-5.5-proPrompts over 272K tokens: $60 input, $270 output. No cached-input price. | OpenAI | $30.00 | $180.00 | Not listed | 1.05M | developers.openai.comchecked 2026-10-10 |
| GPT-5.6 Lunagpt-5.6-lunaPrompts over 272K tokens: $0.40 input, $1.80 output. | OpenAI | $0.20 | $1.20 | $0.02 | 1.05M | developers.openai.comchecked 2026-10-10 |
| GPT-5.6 Solgpt-5.6-solPromotional price, listed as available at least through November 21, 2026. Prompts over 272K tokens: $8 input, $30 output. | OpenAI | $4.00 | $20.00 | $0.40 | 1.05M | developers.openai.comchecked 2026-10-10 |
| GPT-5.6 Terragpt-5.6-terraPrompts over 272K tokens: $4 input, $18 output. | OpenAI | $2.00 | $12.00 | $0.20 | 1.05M | developers.openai.comchecked 2026-10-10 |
| GPT-6 Astragpt-6-astraPrompts over 272K tokens: $20 input, $75 output. | OpenAI | $10.00 | $50.00 | $1.00 | 1.05M | developers.openai.comchecked 2026-10-10 |
| GPT-6 Lunagpt-6-lunaPrompts over 272K tokens: $0.20 input, $0.75 output. | OpenAI | $0.10 | $0.50 | $0.01 | 1.05M | developers.openai.comchecked 2026-10-10 |
| GPT-6 Solgpt-6-solPrompts over 272K tokens: $4 input, $15 output. | OpenAI | $2.00 | $10.00 | $0.20 | 1.05M | developers.openai.comchecked 2026-10-10 |
| GPT-6.1 Solgpt-6.1-solPrompts over 272K tokens: $4 input, $15 output. | OpenAI | $2.00 | $10.00 | $0.10 | 1.05M | developers.openai.comchecked 2026-10-10 |
| Sonarperplexity/sonarSold through Perplexity's Agent API. Each web search costs $0.0025 on top. | Perplexity | $0.25 | $2.50 | $0.0625 | Not listed | docs.perplexity.aichecked 2026-10-10 |
| Grok 4.20grok-4.20-0309-reasoningReasoning version. Prompts of 200K tokens or more: $2.50 input, $5.00 output. | xAI | $1.25 | $2.50 | $0.20 | 1M | docs.x.aichecked 2026-10-10 |
| Grok 4.3grok-4.3Prompts of 200K tokens or more: $2.50 input, $5.00 output. | xAI | $1.25 | $2.50 | $0.20 | 1M | docs.x.aichecked 2026-10-10 |
| Grok 4.5grok-4.5Prompts of 200K tokens or more: $4.00 input, $12.00 output. | xAI | $2.00 | $6.00 | $0.30 | 500K | docs.x.aichecked 2026-10-10 |
| Grok 4.6grok-4.6Prompts of 200K tokens or more: $4.00 input, $12.00 output. | xAI | $2.00 | $6.00 | $0.50 | 500K | docs.x.aichecked 2026-10-10 |
| Grok 4.7grok-4.7Prompts of 200K tokens or more: $4.00 input, $12.00 output. | xAI | $2.00 | $6.00 | $0.50 | 500K | docs.x.aichecked 2026-10-10 |
| Grok Build 0.1grok-build-0.1Model for code. Prompts of 200K tokens or more: $2.00 input, $4.00 output. | xAI | $1.00 | $2.00 | $0.20 | 256K | docs.x.aichecked 2026-10-10 |
| GLM-5.2glm-5.2Storing cached input is free for a limited time. | Zhipu (Z.ai) | $1.40 | $4.40 | $0.26 | 1M | docs.z.aichecked 2026-10-10 |
| GLM-5.3glm-5.3Storing cached input is free for a limited time. | Zhipu (Z.ai) | $1.40 | $4.40 | $0.26 | 1M | docs.z.aichecked 2026-10-10 |
| GLM-5.3-Flashglm-5.3-flashStoring cached input is free for a limited time. | Zhipu (Z.ai) | $0.15 | $0.50 | $0.03 | 1M | docs.z.aichecked 2026-10-10 |
| GLM-5.3-FlashXglm-5.3-flashxStoring cached input is free for a limited time. | Zhipu (Z.ai) | $0.37 | $1.25 | $0.075 | 1M | docs.z.aichecked 2026-10-10 |
Prices change often; we re-check them every 30 days and mark older ones “Needs re-check”. Check the provider's page before you build on a price. Found one out of date? Report a correction.