WHISPER API
Whisper inference, no GPU required.
Send audio, get a transcript, with no GPU to manage. Whisper Large v3 and V3 Turbo behind an OpenAI-compatible endpoint, so existing transcription calls keep working.
Hand this to your agent
Add a Runcrate model to this project: run curl -fsSL"https://runcrate.ai/agents.md?intent=build"and follow it.
Your agent asks for an API key. Create a key
Quick start
Hand it to your agent. Or read the request it sends.
Your agent writes and runs it. The same key works for every model and every GPU.
Available models
Models you can call today. Billed prices, as charged.
| Model | Maker | Price | Detail |
|---|---|---|---|
| openai/whisper-large-v3 | OpenAI | $0.000473 per minute | Full model, for critical transcription |
| openai/whisper-large-v3-turbo | OpenAI | $0.00021 per minute | 8x faster inference, ideal for real-time |
Why Runcrate
Built for production. One key, one balance.
No GPU management
Skip the CUDA setup, model loading, and GPU provisioning. Send a file, receive text. Runcrate handles the infrastructure.
Turbo mode
Whisper Large v3 Turbo delivers results 8x faster than the full model with near-identical word error rates. Pay less, get results faster.
Automatic language detection
Whisper auto-detects the spoken language from 100+ options. No need to specify the language parameter for most use cases.
Parallel requests
Send files in parallel. Each request is billed on its own at the model's listed per-minute price.
FAQ
Common questions.
Related