Cohere Transcribe
Cohere Transcribe is an open-source automatic speech recognition model for highly accurate audio transcription. Cohere says it is built for practical enterprise use, supports 14 languages, uses a 2B parameter Conformer-based encoder-decoder architecture, and currently ranks #1 on the Open ASR Leaderboard for accuracy.
- Type
- Audio
- Weights
- Open
- Downloads
- 212k
- Likes
- 1.2k
- Pipeline
- automatic speech recognition
- License
- apache-2.0
Model hub
CohereLabs/cohere-transcribe-03-2026
- Repo
- CohereLabs/cohere-transcribe-03-2026
- Library
- transformers
- Access
- Gated
Technical tags and languages
More from Cohere
11
-
44k DL
Cohere Transcribe Arabic
Audio
Cohere Transcribe Arabic is a 2B parameter conformer based encoder decoder automatic speech recognition model…
-
Command
Text
AI model by Cohere
-
Command A+ W4A4
Multimodal
Command A+ 05-2026 W4A4 is Cohere’s open-source quantized vision-language reasoning model for agentic, multil…
-
9.5k DL
Command R7B
Multimodal
Command R7B is Cohere's smallest and fastest model in the Command R family, with a 128K context window and 7B…
-
Embed 4
Text
Multimodal embedding model aimed at enterprise search and retrieval over multimodal data.
-
169k DL
North Micro Vision Instruct
Multimodal
North Micro Vision Instruct is a compact 2.4B-parameter open-weight vision-language model from Cohere Labs, p…
-
12k DL
North Mini Code 1.0
Coding
North Mini Code 1.0 is Cohere’s Apache 2.0 agentic coding model with 30B total parameters and 3B active param…
-
106 DL
North Small Translate 1.0
Text
A sparse mixture of experts machine translation model with 25 billion active and 218 billion total parameters…
About Cohere
1
Cohere develops security-focused foundation models and enterprise AI products that can run in cloud, private-cloud, and on-premises environments.