CoolFace
Modelpublic

aufklarer/Cohere-Transcribe-2B-MLX-FP16

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face
0likes24downloads
Model Card

Cohere Transcribe 2B — MLX FP16

MLX weights for Cohere Transcribe 2B, ready for local speech-to-text with speech-swift.

ParametersPrecisionSample rateLanguages
2BFP1616 kHz14

Files

FileDescription
model*.safetensorsMLX model weights
config.jsonModel and runtime configuration
tokenizer.model and tokenizer JSON filesSentencePiece tokenizer data

Usage

bash
speech transcribe recording.wav --engine cohere --model fp16
swift
import CohereTranscribeASR

let model = try await CohereTranscribeModel.load(
    "aufklarer/Cohere-Transcribe-2B-MLX-FP16"
)
let text = model.transcribe(
    audio: samples,
    sampleRate: sourceSampleRate,
    language: "en"
)

Benchmark

English FLEURS read speech, 80 utterances (759.56 seconds), Apple M5 Pro:

VariantBundleWERMean RTFPeak RSS
FP163.85 GiB6.178%0.04443,964 MiB
INT51.62 GiB6.288%0.01501,776 MiB
INT82.25 GiB6.178%0.01592,428 MiB

These read-speech measurements are not Artificial Analysis Conversational or ElevenLabs Scribe results.

Source and license

The base model is CohereLabs/cohere-transcribe-03-2026 and is released under Apache 2.0.

Links