Cohere Transcribe stands out when multilingual speech tests need clean transcripts, stable punctuation, and fewer drops across English, Spanish, and French.
Cohere Transcribe: what stands out
Model link
coherelabs/cohere-transcribe-03-2026 on Wiro
Cover art (generated)

What Cohere Transcribe does
Cohere Transcribe converts speech to text. The key input is the audio file plus the language setting (it does not claim to auto-detect language). The tests below include three short clips generated with text-to-speech plus one sample clip.
Test setup
| Transcribe model | coherelabs/cohere-transcribe-03-2026 |
| maxNewTokens | 256 |
| Audio sources | 3 short TTS clips + 1 sample clip |
5 transcription tests (with audio)
1) English short clip (language = en)
Audio:
Expected text (TTS input): This is a short test sentence for speech to text. It includes numbers like 42 and 3.14, and a name: Jordan.
Transcript:
This is a short test sentence for speech to text. It includes numbers like 42 and 314 and a name, Jordan.
2) Same English clip (language = es)
Audio:
Transcript:
This is a short test sentence for speech to text. It includes numbers like 42 and 314 and a name, Jordan.
3) Spanish short clip (language = es)
Audio:
Expected text (TTS input): Hola, esta es una prueba corta de transcripcion. Incluye el numero cuarenta y dos y la fecha diecisiete de abril.
Transcript:
Hola, esta es una prueba corta de transcripción. Incluye el número 42 y la fecha 17 de abril.
4) French short clip (language = fr)
Audio:
Expected text (TTS input): Bonjour, ceci est un court test de transcription. Il contient le nombre quarante-deux et la date dix-sept avril.
Transcript:
Bonjour. Ceci est un court test de transcription. Il contient le nombre 42 et la date 17 avril.
5) Sample clip (language = en)
Audio:
Transcript:
Finally, there are many small cats including loose pet cats that eat the far more numerous small prey like insects, rodents, lizards and birds.
Quick takeaways
- Short clips transcribe cleanly, including punctuation in these samples.
- Numbers often normalize to digits (for example 42, 17).
- Decimal formatting can vary (3.14 became 314 in this run).
Try it
Run the model here: coherelabs/cohere-transcribe-03-2026