CohereLabs/cohere-transcribe-arabic-07-2026 · Hugging Face

Source: https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026 Cohere Transcribe Arabic is an open source release of a 2B parameter dedicated audio-in, text-out automatic speech recognition (ASR) model. The optimized for Arabic and English, with a focus on Arabic dialect performance and Arabic-English code-switching. Based on theCohere Transcribearchitecture.

Developed by:CohereandCohere Labs. Point of Contact:Cohere Labs.

Namecohere-transcribe-arabic-07-2026Architectureconformer-based encoder-decoderInputaudio waveform → log-Mel spectrogram. Audio is automatically resampled to 16kHz if necessary during preprocessing. Similarly, multi-channel (stereo) inputs are averaged to produce a single channel signal.Outputtranscribed textModela large Conformer encoder extracts acoustic representations, followed by a lightweight Transformer decoder for token generationTraining objectivesupervised cross-entropy on output tokensLanguages- Arabic

• English

LicenseApache 2.0✨Try the Cohere Transcribe Arabicdemo✨

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#usageUsage

Cohere Transcribe Arabic is supported natively intransformers. This is the recommended way to use the model for offline inference. For online inference, see the vLLM integration example below.

pip install transformers>=5.4.0 torch huggingface_hub soundfile librosa sentencepiece protobuf accelerate

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#quick-start-%F0%9F%A4%97Quick Start 🤗

Transcribe any audio file in a few lines:

from transformers import AutoProcessor, CohereAsrForConditionalGeneration
from transformers.audio_utils import load_audio
from huggingface_hub import hf_hub_download

processor = AutoProcessor.from_pretrained("CohereLabs/cohere-transcribe-arabic-07-2026")
model = CohereAsrForConditionalGeneration.from_pretrained("CohereLabs/cohere-transcribe-arabic-07-2026", device_map="auto")

# Example: transcribe Arabic audio
audio_file = "your_audio.wav"
audio = load_audio(audio_file, sampling_rate=16000)

inputs = processor(audio, sampling_rate=16000, return_tensors="pt", language="ar")
inputs.to(model.device, dtype=model.dtype)

outputs = model.generate(**inputs, max_new_tokens=256)
text = processor.decode(outputs, skip_special_tokens=True)
print(text)

Long-form transcriptionFor audio longer than the feature extractor’smax\_audio\_clip\_s, the feature extractor automatically splits the waveform into chunks. The processor reassembles the per-chunk transcriptions using the returnedaudio\_chunk\_index.

from transformers import AutoProcessor, CohereAsrForConditionalGeneration
import time

processor = AutoProcessor.from_pretrained("CohereLabs/cohere-transcribe-arabic-07-2026")
model = CohereAsrForConditionalGeneration.from_pretrained("CohereLabs/cohere-transcribe-arabic-07-2026", device_map="auto")

audio = load_audio("your_long_audio.wav", sampling_rate=16000)
sr = 16000
duration_s = len(audio) / sr
print(f"Audio duration: {duration_s / 60:.1f} minutes")

inputs = processor(audio=audio, sampling_rate=sr, return_tensors="pt", language="ar")
audio_chunk_index = inputs.get("audio_chunk_index")
inputs.to(model.device, dtype=model.dtype)

start = time.time()
outputs = model.generate(**inputs, max_new_tokens=256)
text = processor.decode(outputs, skip_special_tokens=True, audio_chunk_index=audio_chunk_index, language="ar")[0]
elapsed = time.time() - start
rtfx = duration_s / elapsed
print(f"Transcribed in {elapsed:.1f}s — RTFx: {rtfx:.1f}")
print(text)

English transcriptionThe model also supports English. Specifylanguage="en":

inputs = processor(audio, sampling_rate=16000, return_tensors="pt", language="en")
inputs.to(model.device, dtype=model.dtype)

outputs = model.generate(**inputs, max_new_tokens=256)
text = processor.decode(outputs, skip_special_tokens=True)
print(text)

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#vllm-integrationvLLM Integration

For production serving we recommend running via vLLM following the instructions below.

Run cohere-transcribe-arabic-07-2026 via vLLMFirst install vLLM (refer tovLLM installation instructions):

uv venv --python 3.12 --seed
source .venv/bin/activate

uv pip install -U vllm==0.19.0 --torch-backend=auto
uv pip install vllm[audio]
uv pip install librosa

Start vLLM server

vllm serve CohereLabs/cohere-transcribe-arabic-07-2026 --trust-remote-code

Send request

curl -v -X POST http://localhost:8000/v1/audio/transcriptions \
 -H "Authorization: Bearer $VLLM_API_KEY" \
-F "file=@$(realpath ${AUDIO_PATH})" \
-F "model=CohereLabs/cohere-transcribe-arabic-07-2026"

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#resultsResults

Open Universal Arabic ASR Leaderboard (as of 07.07.2026)ModelAverageWER · CERSADAWER · CERCommon VoiceWER · CERMASC cleanWER · CERMASC noisyWER · CERMGB-2WER · CERCasablancaWER · CERCohere Transcribe Arabic 07-202625.8711.8037.4723.535.821.6219.606.4527.0710.1315.548.4049.71****20.66OmniASR LLM 7B28.3212.5241.6124.958.752.7119.695.7629.2910.6614.137.1056.4623.96OmniASR LLM 3B29.9613.7746.1827.279.152.8019.906.1330.0311.2714.227.0660.2728.06OmniASR LLM 1B29.9613.4043.8424.549.552.9720.036.1430.2611.1815.347.5660.6828.02Cohere Transcribe 03-202630.6716.3760.1145.448.172.498.662.9719.01****7.7125.339.2862.7130.31Qwen3-Omni 30B30.7113.6744.8226.1111.464.2821.475.5930.8511.2813.09****6.2062.5528.53NVIDIA Conformer-CTC (LM)32.9113.8444.5223.768.802.7723.745.6334.2911.0717.206.8768.9032.97OmniASR LLM 300M32.9614.8451.3829.1012.034.0420.666.2232.4512.2316.587.8664.6429.61Gemma 4 E4B32.9813.7143.4020.9619.657.4824.867.7633.5912.2517.728.6758.6325.11Qwen3-ASR 1.7B33.3612.3345.5319.9016.905.0624.375.7234.2910.8416.576.2564.4726.23Voxtral-Small 24B34.4715.2950.8228.8515.255.5423.967.0634.4312.2216.037.4166.3030.64NVIDIA Conformer-CTC (greedy)34.7413.3747.2622.5410.603.0524.125.6335.6411.0219.697.4671.1330.50Gemma 4 E2B35.8715.3446.2323.4723.769.1327.478.9936.1513.9320.7210.1560.8726.35Whisper Large v336.8617.2155.9634.6217.835.7424.667.2434.6312.8916.267.7471.8135.04 Link to the live leaderboard:Open Universal Arabic ASR Leaderboard.

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#resourcesResources

For more details and results:

• Technical blog postcontains WERs and other quality metrics.
• Announcement blog postfor more information about the model.
• TheOpen Universal Arabic ASR Leaderboard.

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#strengths-and-limitationsStrengths and Limitations

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#strengthsStrengths

Cohere Transcribe Arabic demonstrates strong transcription accuracy for Arabic and English. As a dedicated speech recognition model, it benefits from efficient inference via the Conformer encoder-decoder architecture.

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#limitationsLimitations

• **Single language.**The model performs best when remaining in-distribution of a single, pre-specified language. It does not feature explicit, automatic language detection and exhibits inconsistent performance on code-switched audio.
• **Timestamps/Speaker diarization.**The model does not feature either of these.
• **Silence.**Like most AED speech models, Cohere Transcibe Arabic is eager to transcribe, even non-speech sounds. The model benefits from prepending a noise gate or VAD (voice activity detection) model in order to prevent low-volume, floor noise from turning into hallucinations.

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#model-card-contactModel Card Contact

For errors or additional questions about details in this model card, contact[email protected]or raise an issue.

Terms of Use: We hope that the release of this model will make community-based research efforts into Arabic speech more accessible. This model is governed by an Apache 2.0 license.

https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026#citationCitation

To cite this model please use the following bibtex:

@misc{shaun_cassini_2026,
    author       = { Shaun Cassini and Sebastian Vincent and Xiaolu Lu and Julian Mack and Dhruti Joshi and Pierre Richemond },
    title        = { cohere-transcribe-arabic-07-2026 (Revision 0a8193c) },
    year         = 2026,
    url          = { https://huggingface.co/CohereLabs/cohere-transcribe-arabic-07-2026 },
    doi          = { 10.57967/hf/9549 },
    publisher    = { Hugging Face }
}