Whisper
OpenAIAUDIO Open weight
- Provider
- OpenAI
- Modality
- AUDIO
- Parameters
- —
- Context window
- —
- Weights
- Open
- Released
- —
Whisper is OpenAI's open speech-to-text model, with strong multilingual transcription including Hindi.
Because it is open and free, it is the default choice for self-hosted transcription and subtitles.
Best for
- Transcription and subtitles
- Multilingual and Hindi audio
- Private, self-hosted speech-to-text
Strengths & how it compares
- Free, open and accurate across many languages.
- Runs locally or via a low-cost hosted API.
- Vs cloud STT services: free and private; large models need a GPU for speed.
How to access
Free (open weights); hosted API per minute.
Access Whisper