AI API prices: GPT, Claude, Gemini, DeepSeek
/ Price listPay per request from your balance: a request costs its tokens at the prices below, and every model works with the same OpenCheese API key in the OpenAI- and Anthropic-compatible APIs. Prefer a fixed monthly price? Compare the plans.
48 models, up to 96% below list prices.
Text models
Prices in USD per 1M tokens. A struck-through price is the list price; the percentage beside a model is its deepest input or output discount.
| Model | Effort | Input | Output | Cache read | Cache write |
|---|---|---|---|---|---|
Claude Fable 5.1 −90%claude-fable-5-1 | low, medium, high, xhigh, max, ultracode | $3.50 | $5.00 | $0.1875 | $5.00 |
Claude Haiku 5.5 −50%claude-haiku-5-5 | low, medium, high, xhigh, max | $0.05 | $0.25 | $0.005 | $0.10 |
Claude Opus 5 −96%claude-opus-5 | low, medium, high, xhigh, max, ultracode | $0.40 | $1.00 | $0.05 | $1.00 |
Claude Opus 5.5 −95%claude-opus-5-5 | low, medium, high, xhigh, max, ultracode | $0.48 | $1.00 | $0.054 | $0.96 |
Claude Sonnet 5.5 −93%claude-sonnet-5-5 | low, medium, high, xhigh, max, ultracode | $0.16 | $0.70 | $0.05 | $0.9375 |
DeepSeek V4 Pro −20%deepseek-v4-pro | low, high, max | $1.392 | $2.784 | $0.0145 | — |
DeepSeek V4.1 Flash −50%deepseek-v4.1-flash | low, high, max | $0.15 | $0.60 | $0.007 | — |
Gemini 3.8 Flash −75%gemini-3.8-flash | low, medium, high | $0.195 | $0.9375 | $0.0975 | $0.75 |
Gemini 4 Argon −90%gemini-4-argon | low, medium, high | $0.30 | $1.00 | $0.05 | $0.20 |
Gemma 4 31B IT −10%gemma-4-31b-it | — | $0.108 | $0.324 | $0.01 | $0.005 |
GLM 5.3 −15%glm-5.3 | low, high, max | $1.19 | $3.96 | $0.247 | — |
GLM 5.3 Flash −50%glm-5.3-flash | low, high, max | $0.075 | $0.25 | $0.03 | — |
GPT 6 Astra −94%gpt-6-astra | low, medium, high, xhigh, max, ultra | $1.00 | $3.00 | $0.14 | $1.25 |
GPT 6 Luna −80%gpt-6-luna | low, medium, high, xhigh, max | $0.06 | $0.10 | $0.01 | $0.125 |
GPT 6 Luna Fast −80%gpt-6-luna-fast | low, medium, high, xhigh, max | $0.12 | $0.20 | $0.02 | $0.25 |
GPT 6.1 Sol −91%gpt-6.1-sol | low, medium, high, xhigh, max, ultra | $0.24 | $0.90 | $0.056 | $1.00 |
GPT 6.1 Sol Fast −91%gpt-6.1-sol-fast | low, medium, high, xhigh, max, ultra | $0.48 | $1.80 | $0.112 | $2.00 |
Grok 4.7 −86%grok-4.7 | low, medium, high, xhigh | $0.28 | $0.84 | $0.125 | — |
Kimi K2.7 Code −20%kimi-k2.7-code | low, medium, high | $0.76 | $3.52 | $0.171 | — |
Kimi K3 −20%kimi-k3 | low, high, max | $2.52 | $12.00 | $0.252 | — |
Mimo V2.6 Flash −5%mimo-v2.6-flash | low, medium, max | $0.147 | $0.285 | $0.0057 | — |
Mimo V2.6 Promimo-v2.6-pro | low, medium, max | $0.46 | $0.90 | $0.006 | — |
MiniMax M3 −50%minimax-m3 | low, medium, max | $0.30 | $1.20 | $0.084 | $0.60 |
MiniMax M3.1 Flashminimax-m3.1-flash | low, high, max | $0.155 | $0.605 | $0.02 | — |
Muse Spark 1.3muse-spark-1.3 | minimal, low, medium, high, xhigh, max | $0.11 | $0.21 | $0.003 | — |
Qwen 3.7 Flash −50%qwen3.7-flash | low, medium, max | $0.07 | $0.20 | $0.018 | $0.0875 |
Qwen 3.7 Plusqwen3.7-plus | low, medium, max | $0.32 | $1.29 | $0.07 | $0.42 |
Qwen 3.8 Flash −5%qwen-3.8-flash | low, medium, max | $0.1425 | $0.4465 | $0.021 | $0.20 |
Qwen 3.8 Maxqwen3.8-max | low, high, max | $2.10 | $6.10 | $0.21 | $2.60 |
Image models
Priced per image
Prices in USD per generated image, by output resolution. Each source image of an edit costs the source image price.
| Model | 768 | 1K | 1.5K | 2K | 4K | Source image |
|---|---|---|---|---|---|---|
FLUX.3 Imageflux-3-image | $0.041 | $0.048 | $0.07 | $0.10 | $0.607 | — |
Grok Imagine 2grok-imagine-image-2.0 | — | low $0.045medium $0.065 | — | low $0.065medium $0.085 | — | $0.015 |
Muse Imagemuse-image | — | $0.015 | $0.015 | $0.015 | $0.015 | $0.0035 |
Qwen Image 3qwen-image-3.0 | — | $0.035 | — | $0.035 | — | $0.0035 |
Qwen Image 3 Proqwen-image-3.0-pro | — | $0.04 | — | $0.075 | — | $0.003 |
Seedream 5.0 Liteseedream-5.0-lite | — | $0.04 | — | $0.04 | $0.04 | — |
Seedream 5.0 Proseedream-5.0-pro | — | $0.045 | — | $0.09 | — | $0.003 |
Wan 2.7 Imagewan2.7-image | — | $0.035 | — | $0.035 | — | $0.0035 |
Wan 2.7 Image Prowan2.7-image-pro | — | $0.08 | — | $0.085 | $0.09 | $0.0035 |
Priced per token
Prices in USD per 1M tokens. A struck-through price is the list price; the percentage beside a model is its deepest input or output discount.
| Model | Input | Output | Cache read | Cache write |
|---|---|---|---|---|
GPT Image 2 −75%gpt-image-2 | $1.25 | $7.50 | $0.0625 | — |
GPT Image 2.5 Flare −75%gpt-image-2.5-flare | $1.25 | $7.50 | $0.0625 | — |
GPT Image 2.5 Sunburst −75%gpt-image-2.5-sunburst | $1.25 | $7.50 | $0.0625 | — |
Nano Banana 2gemini-3.1-flash-image | $0.50 | $60.00 | — | — |
Nano Banana 2 Litegemini-3.1-flash-lite-image | $0.25 | $30.00 | — | — |
Nano Banana 2.1gemini-nano-banana-2.1 | $1.50 | $30.00 | — | — |
Nano Banana Progemini-3-pro-image | $2.00 | $120.00 | — | — |
Free models
Free models cost nothing per request. They unlock once your balance reaches the amount shown, so one small top-up is enough.
| Model | Effort | Price | Unlocks at balance |
|---|---|---|---|
GPT OSS 120Bgpt-oss-120b | — | Free | $0.11 |
Qwen 3.6 35B FP8Qwen3.6-35B-A3B-FP8 | — | Free | $0.11 |
Qwen 3.8 27B FP8Qwen3.8-27B-Q8 | — | Free | $0.11 |
Top-up bonus
Every top-up adds extra balance on top of what you pay. The bonus grows with the amount, rising smoothly between these steps. Top-ups start at $5.
- $10+2%
- $20+4%
- $40+6%
- $80+8%
- $160+16%
- $320+24%
- $640+40%
- $1,280+64%
- $2,560+100%
Prices are in US dollars and come straight from the gateway, so this page always shows the current ones. Some models cost more for very long prompts; after signing in, the Prices page in your account shows every price, those included.
Sell us your unused API credits
Have OpenAI, Anthropic (Claude), Google Gemini or DeepSeek credits you won't use before they expire, a key with a balance, spare model capacity or GPU servers? We buy them; anything that relies on a provider account needs the provider's written permission. Unlike cheese, credits don't get better with age.