Pricing
Pay-as-you-go. No subscriptions, no minimums, no surprises.
How it works
- ·You only pay for the model calls your code actually makes — tokens for text models, seconds of output for video, per generation for image.
- ·Every call writes a usage row you can audit in the dashboard — which model, what inputs, what it cost.
- ·Top up your project balance once and it's drawn down as jobs run. When it hits zero, jobs pause until you refill.
A few examples
These are what your project gets billed per call.
GPT-4o
openai/gpt-4o
$2.63 / $10.50
per 1M input / output tokens
Gemini 2.5 Pro
google/gemini-2-5-pro
$1.31 / $10.50
per 1M input / output tokens
GPT Image 2
openai/gpt-image-2
from $0.222
per generation
- low · 1024x768$0.0053/image
- low · 1024x1024$0.0063/image
- low · 1024x1536$0.0053/image
- low · 1920x1080$0.0053/image
- low · 2560x1440$0.0074/image
- low · 3840x2160$0.013/image
- medium · 1024x768$0.039/image
- medium · 1024x1024$0.056/image
- medium · 1024x1536$0.044/image
- medium · 1920x1080$0.042/image
- medium · 2560x1440$0.059/image
- medium · 3840x2160$0.106/image
- high · 1024x768$0.152/image
- high · 1024x1024$0.222/image
- high · 1024x1536$0.173/image
- high · 1920x1080$0.166/image
- high · 2560x1440$0.233/image
- high · 3840x2160$0.421/image
Seedream v4 (edit)
bytedance/seedream-v4-edit
$0.032
per generation
Kling v3 Pro — image→video
kuaishou/kling-v3-i2v
from $0.12
per second of output
- audio off$0.118/s
- audio on$0.176/s
- audio + voice control$0.206/s
Plus the rest of the OpenAI, Gemini, Kling, and Seedance lineups — all billed the same way.
Get started
Sign in, create a project, top up, and start shipping.
Sign in with Google