Whisper
Whisper is OpenAI's speech-to-text model available through its API that has become a widely-used baseline for automatic speech recognition tasks. Recent coverage shows Whisper competing with newer open-source ASR models in accuracy benchmarks, while also being optimized through techniques like speculative decoding for faster inference and integrated into production systems via platforms like Hugging Face and Together AI for real-time voice applications.
Updated 3 August 2026
Specifications
No specifications recorded yet.
Latest developments
How to become a 10x ramble-coder
TLDR · 1 week ago ·
9
Best Open Speech Recognition (ASR) Models in 2026: WER, Languages, Latency, and License Compared
MarkTechPost · 1 week ago ·
20
How speech models fail where it matters the most and what to do about it
Together AI · 5 months ago ·
39
Next generation medical image interpretation with MedGemma 1.5 and medical speech to text with MedASR
Google Research · 6 months ago ·
38
Announcing the fastest inference for realtime voice AI agents
Together AI · 8 months ago ·
50
2026
- How to become a 10x ramble-coder
- Best Open Speech Recognition (ASR) Models in 2026: WER, Languages, Latency, and License Compared
- How speech models fail where it matters the most and what to do about it
- Next generation medical image interpretation with MedGemma 1.5 and medical speech to text with MedASR
2025
- Announcing the fastest inference for realtime voice AI agents
- Voxtral
- Blazingly fast whisper transcriptions with Inference Endpoints
2024
- Powerful ASR + diarization + speculative decoding with Hugging Face Inference Endpoints
- GPT-4 API general availability and deprecation of older models in the Completions API
- Fine-Tune W2V2-Bert for low-resource ASR with 🤗 Transformers
2023
Relationships
Products & technology
- OpenAI develops this model · 3 sources
- Together AI deploys this model · 1 source
- Together AI integrated with this model · 1 source
- Hugging Face integrated with this model · 1 source
- Hugging Face deploys this model · 1 source