Skip to main content
Prices are in USD and charged from your prepaid balance. Image models cost a flat price per image, set by the size tier. The quality setting does not change the price. Failed images are free.

​
Image models

A dash means the tier is not offered for that model.

​
Which model to pick

  • Fast drafts and high volume: gpt-image-2.5-flare (the fast GPT Image model) or nano-banana-2.1
  • Final, high-quality images: gpt-image-2.5-sunburst (the quality GPT Image model) or nano-banana-pro
  • Lowest price at 1K: nano-banana or nano-banana-2-lite, then gpt-image-2.5-flare

​
Video models

Price per second of video. A job costs the per-second rate times duration. See the Video API. Seedance 2.5 makes 4 to 30 second videos. Seedance 2.0 makes 5 to 15 second videos and is the only one with 4K.

​
Text model

POST /v1/chat/completions with a standard OpenAI chat body. Model ID: gpt-6.1-sol. Prices per 1M tokens, by request context size:

​
Your live price list

Sign in and open Models to see the current list and your prices. Your key can also list the models it may use:
If a price here differs from your contract, your contract applies. Ask your manager.