Model pricing
Text models are priced per 1M tokens, image models per image, and video models per generated second. The last column compares our price with the official reference.
Text modelsUSD per 1M tokens
| Model | Input | Output | Cache read | Cache write | Vs official |
|---|---|---|---|---|---|
| Claude | |||||
| claude-fable-5 | $4 | $20 | $0.4 | $5 | 0.4× |
| claude-fable-5-1 | $4 | $20 | $0.1 | $5 | 0.4× |
| claude-haiku-5-5 | $0.04 | $0.2 | $0.004 | $0.05 | — |
| claude-opus-4-6 | $2 | $10 | $0.2 | $2.5 | 0.4× |
| claude-opus-4-7 | $2 | $10 | $0.2 | $2.5 | 0.4× |
| claude-opus-4-8 | $2 | $10 | $0.2 | $2.5 | 0.4× |
| claude-opus-5 | $2 | $10 | $0.2 | $2.5 | 0.4× |
| claude-opus-5-5 | $1.6 | $8 | $0.08 | $2 | 0.4× |
| claude-sonnet-4-6 | $1.2 | $6 | $0.12 | $1.5 | 0.4× |
| claude-sonnet-5 | $0.8 | $4 | $0.08 | $1 | 0.4× |
| claude-sonnet-5-5 | $0.8 | $4 | $0.08 | $1 | 0.4× |
| GPT | |||||
| gpt-5.5 | $1.5 | $9 | $0.15 | $0 | 0.3× |
| gpt-5.6-sol | $1.2 | $6 | $0.12 | $1.5 | 0.3× |
| gpt-5.6-terra | $0.6 | $3.6 | $0.06 | $0.75 | 0.3× |
| gpt-6-astra | $3 | $15 | $0.3 | $3.75 | 0.3× |
| gpt-6-sol | $0.6 | $3 | $0.06 | $0.75 | 0.3× |
| gpt-6.1-sol | $0.6 | $3 | $0.06 | $0.75 | 0.3×–0.6× |
| Grok | |||||
| grok-4.5 | $0.6 | $1.8 | $0.09 | $0 | 0.3× |
| grok-4.6 | $0.6 | $1.8 | $0.15 | $0 | 0.3× |
| grok-4.7 | $0.6 | $1.8 | $0.15 | $0 | 0.3× |
| grok-4.7-build-fast | $1.2 | $3.6 | $0.3 | $0 | 0.3× |
| DeepSeek | |||||
| deepseek-v4-pro | $0.858 | $2.57 | $0.0286 | $0 | 0.65× |
| deepseek-v4.1-flash | $0.24 | $0.96 | $0.0048 | — | 0.82×* |
| GLM | |||||
| glm-5.2 | $0.91 | $2.86 | $0.169 | $0 | 0.65× |
| glm-5.3 | $1 | $3.2 | $0.22 | — | 0.78×–0.85×* |
| Gemini | |||||
| gemini-3.1-pro | $1 | $6 | $0.1 | — | 0.5×* |
| gemini-3.7-flash | $0.375 | $1.87 | $0.0375 | — | 0.5×* |
| gemini-3.8-flash | $0.375 | $1.87 | $0.0375 | — | 0.5×* |
| Kimi | |||||
| kimi-k3 | $2.2 | $11 | $0.22 | — | 0.74×* |
| Other | |||||
| qwen3.7-plus | $0.2 | $0.8 | $0.04 | $0.25 | 0.68×* |
| qwen3.8-27b | $0.3 | $1.2 | $0.06 | — | 0.68×* |
| qwen3.8-flash | $0.08 | $0.27 | $0.01 | $0.125 | 0.68×* |
Image modelsUSD per generated image, any size or quality
| Model | Per image | Vs official |
|---|---|---|
| gpt-image-2 | $0.15 / image | 0.72× |
| gpt-image-2.5 | $0.15 / image | — |
| gpt-image-2.5-flare | $0.3 / image | — |
| gpt-image-2.5-sunburst | $0.35 / image | — |
Video modelsUSD per generated second, by output resolution
| Model | 480p | 720p | 1080p | Input image | Vs official |
|---|---|---|---|---|---|
| grok-imagine-video-1.5 | $0.08 / second | $0.14 / second | $0.25 / second | $0.01 / image | 1× |
| wan3.0-video Billed on generated and reference video seconds. | $0.033 / second | $0.066 / second | $0.132 / second | — | 0.66× |
Prices are shown in USD credit. A dash means the model has no separate cache price. A multiplier range means input, output or cache discounts differ, and an asterisk marks an incomplete comparison. Your signed-in account determines which models are available, and each completed request appears in Usage.