Whisper
Free tierOpenAI's open-source speech recognition — best-in-class accuracy across 99 languages.
Free·Technical·Powered by OpenAI·API available·Open source
Key strengths
Multilingual speech recognition and transcriptionTranslation from multiple languages into EnglishRobustness to accents, background noise, and technical languageTrained on 680,000 hours of multilingual and multitask supervised data50% fewer errors than specialized models in zero-shot evaluationOpen-source models and inference code (MIT License)
Completely free
San Francisco, US
Founded 2022
Self-hostable
Facts verified against openai.com on
Get alerts when Whisper changes
Free. We'll email you when pricing, the free tier, API access or open-source status changes.
No ratings yet
Whisper — Technical Overview
Whisper is an open-source automatic speech recognition (ASR) system developed by OpenAI, released in 2022 under the MIT License. It is trained on 680,000 hours of multilingual and multitask weakly supervised data, enabling strong zero-shot generalization across languages, accents, and acoustic environments.
Key technical characteristics:
- Multitask architecture: Supports transcription, translation (into English), language identification, and voice activity detection within a single model.
- Multilingual support: Recognizes and transcribes speech across a wide range of languages, and translates from multiple languages into English.
- Robustness: Demonstrates resilience to accents, background noise, and technical or domain-specific vocabulary.
- Zero-shot performance: Achieves ~50% fewer errors than comparable specialized models in zero-shot evaluation settings.
- Open weights & inference code: Model weights and inference code are publicly available on GitHub (110,000+ stars) under the MIT License, making the system fully self-hostable.
- Pricing: Free to use; no API cost for self-hosted deployments.
Alternatives
See all alternatives Alternatives to Whisper
Is Whisper yours? Claim the listing & get the badge