Pricing and Billing
Inference Pricing
Inference usage is calculated based on different units for each model. See the tables below for pricing based on model type.
Language Models
| Model | Price |
|---|---|
Llama-2-13B-chat | $0.24 / 1M tokens |
Llama-2-70B-chat | $0.99 / 1M tokens |
Falcon-40B-instruct | $0.66 / 1M tokens |
Image Models
| Model | Resolution | Price |
|---|---|---|
Stable Diffusion XL | all resolutions | $0.0024 / image |
Upscaling Models
| Model | Price |
|---|---|
Real-ESRGAN | $0.003 / image |
GFPGAN | $0.004 / image |
Audio Models
| Model | Mode | Price |
|---|---|---|
Whisper-large-v2 or v3 | transcription or translation | $0.0010 / minute of audio |
Whisper-large-v2 or v3 | transcription or translation + alignment and/or diarization | $0.0020 / minute of audio |
Making Payments
You will receive a monthly invoice to your account email address for inference usage. Feel free to contact support with questions.
We are working hard to improve metrics and payment systems for our inference services. Please bear with us while we grow!
Last updated