Pricing
API
Pay per output and per token from a prepaid dollar balance. No subscription, no seats, no minimum.
Video
- Generation
- from 1.13¢ / second
- Avatar & lip-sync
- from 2.5¢ / second
- Upscaling & restoration
- from 2.14¢ / second
- Editing & motion transfer
- from 7¢ / second
- Relight & restyle
- from 20¢ / second
- Background removal
- from 2.25¢ / 30 frames
Image
- Generation
- from 0.59¢ / generation
- Generation per megapixel
- from 0.5¢ / MP
- Editing
- from 0.59¢ / generation
- Upscaling
- from 8¢ / 24 MP
Audio
- Speech
- from 5¢ / 1,000 characters
- Voice conversion
- from 16.8¢ / minute
- Music
- from 60¢ / minute
- Sound effects
- from 0.71¢ / second
- Isolation & cleanup
- from 10¢ / minute
- Voice cloning
- free
LLMs & VLMs
Text generation, reasoning, and image understanding, billed per token. Reasoning tokens count as output.
- Input tokens
- from $0.15 / 1M tokens
- Cached input
- from $0.007 / 1M tokens
- Output tokens
- from $0.50 / 1M tokens
Full rates and billing units are listed for each model in the catalog. Estimate media jobs before submitting; LLM and VLM charges follow actual token usage.
Get more from Hedra Inference and Research
Dedicated serving, model training, and GPU execution are priced per engagement, against your model, hardware, and workload.