Whisper medium vs Whisper turbo
Two large-ish Whisper checkpoints with very different speed profiles. Whisper turbo ranks higher in 4 of the matcher's seven common setups (with 3 tied), but the right choice still depends on the differences below.
Key differences
- Only Whisper medium supports speech translation.
- Whisper turbo carries the higher relative speed rating (5/5 versus 2/5).
- OpenAI lists about 5 GB of VRAM for Whisper medium and about 6 GB for Whisper turbo.
Side by side
| Fact | Whisper medium | Whisper turbo |
|---|---|---|
| Provider · size | OpenAI · 769M | OpenAI · 809M |
| License | MIT Permissive | MIT Permissive |
| Languages | 99+ listed languages | 99+ listed languages |
| Streaming | Via runtime | Via runtime |
| Timestamps | Yes | Yes |
| Speaker diarization | Separate pipeline | Separate pipeline |
| CPU · Apple · NVIDIA fit | 2/5 · 4/5 · 5/5 | 2/5 · 4/5 · 5/5 |
| Live · long recordings | 2/5 · 5/5 | 3/5 · 5/5 |
| Main runtime | openai-whisper | openai-whisper |
Which ranks higher in common setups
Scores from the same deterministic matcher as the homepage, applied to all 26 reviewed models (rank out of 26). They are decision guidance, not accuracy claims.
| Setup | Whisper medium | Whisper turbo | Higher |
|---|---|---|---|
| Long recordings on an 8–12 GB NVIDIA GPU, balanced | 76/100 · #5 | 79/100 · #1 | Whisper turbo |
| Live captions or voice input (native streaming + timestamps, speed first) | 70/100 · #16 | 81/100 · #7 | Whisper turbo |
| Subtitles with timing, accuracy first | 87/100 · #1 | 87/100 · #2 | Tie |
| Meetings with speaker labels (diarization + timestamps) | 84/100 · #2 | 84/100 · #3 | Tie |
| CPU-only machine, low memory use | 59/100 · #14 | 59/100 · #15 | Tie |
| Apple Silicon with 8–16 GB unified memory, long recordings | 66/100 · #14 | 69/100 · #8 | Whisper turbo |
| Japanese live transcription (native streaming, speed first) | 64/100 · #12 | 76/100 · #6 | Whisper turbo |
Language coverage
Both list 100 languages: Afrikaans, Albanian, Amharic, Arabic, Armenian, Assamese, Azerbaijani, Bangla, Bashkir, Basque, Belarusian, Bosnian and 88 more. Whisper medium lists no language that Whisper turbo lacks. Whisper turbo lists no language that Whisper medium lacks.
How to run each
- Whisper medium
- openai-whisper, Hugging Face Transformers, whisper.cpp, faster-whisper.
pip install -U openai-whisperOpenAI lists about 5 GB of VRAM for medium (769M parameters); whisper.cpp lists about 2.1 GB of memory.
- Whisper turbo
- openai-whisper, Hugging Face Transformers, whisper.cpp, faster-whisper.
pip install -U openai-whisperOpenAI lists about 6 GB of VRAM for turbo (809M parameters), an optimized version of large-v3. OpenAI also notes that turbo is not trained for translation tasks.
Runtimes and figures from the official sources checked on November 2, 2026; details on each Whisper medium and Whisper turbo profile.
Common questions
Is Whisper medium better than Whisper turbo?
It depends on the job. Whisper turbo ranks higher in 4 of the matcher's seven common setups (with 3 tied), but the right choice still depends on the differences below. Benchmark both on your own audio before deciding.
Which is lighter to run, Whisper medium or Whisper turbo?
They carry the same memory rating (2/5). See the official guidance quoted in the run section above.
Which supports more languages?
Both list 100 languages in our record.
Go further
- Add a third model in the interactive comparison
- Benchmark both on your own audio
- All model comparisons
Generated from the reviewed model database (latest review November 3, 2026). Differences are taken from primary sources; matcher scores are Speech Model Match's editorial decision scores.