Project AURIONAI Use-Cases Research & Innovation
speech recognition

Cohere Transcribe (cohere-transcribe-03-2026)

Cohere open ASR model; led the Open ASR Leaderboard in March 2026.

CohereCohereCanada
Cohere
Levelmodel
Classspeech recognition
Parameters2B
Contextnot in the notes
Released26 Mar 2026
Opennessopen weights permissive
Runs onsingle datacentre GPU
Statusactive

Benchmark Results

vendor-reportedmaintainer-run
Hugging Face Open ASR Leaderboardspeech
eight-set macro-average English WERmaintainer-run
5.42
048

Score in WER. Lower is better.

rank 2, snapshot 2026-05-06 via benchmarklist6 May 2026 benchmarklist.com
LibriSpeechspeech
test-othervendor-reported
2.37
048

Score in WER. Lower is better.

Mar 2026 select.keenable.ai
test-cleanvendor-reported
1.25
048

Score in WER. Lower is better.

Mar 2026 select.keenable.ai
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).

R&D Topics

Massively multilingual speech recognitionproduced by

Identity

Developer
Cohere
Family
not in the notes
Released
26 Mar 2026 sourcessecondary report
Weights published
–
Record checked
10 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
autoregressive Conformer + LLM decoder (pattern, unverified for this model)
Total parameters
2B sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
audio
Output
text
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights permissive
Licence
Apache License 2.0 sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
single datacentre GPU
Footprint
not in the notes
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Lineage

The notes record no lineage for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. Speech Recognition Accuracy Leaders, 2022–2026 select.keenable.ai https://select.keenable.ai/r/speech-recognition-accuracy-leaders-2022-2026/JbURQjlVooLQ6J10NkbRVw
  2. Open ASR Leaderboard Benchmark Scores & AI Model Leaderboard | BenchmarkList benchmarklist.com https://benchmarklist.com/benchmarks/open_asr_leaderboard/

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Cohere Transcribe (Cohere-Transcribe-03-2026)", https://aurion.semai.id/models/cohere-transcribe/, records of October 2026.