Qwen3-ASR-0.6B

Name: Qwen3-ASR-0.6B
Author: Qwen

by Qwen

Ultra-fast 0.6B ASR model supporting 52 languages with 92ms time-to-first-token

2.9Mdl/month

341likes

938Mparams

HuggingFace Run it on your own data, free

Identifiers

Model ID

Qwen/Qwen3-ASR-0.6B

Feature URI

mixpeek://transcription@v1/qwen3_asr_06b_v1

Overview

Qwen3-ASR-0.6B is a compact automatic speech recognition model supporting 52 languages and dialects. It achieves 92ms average time-to-first-token, making it suitable for real-time transcription. At high concurrency (128), it can process 2000 seconds of speech per second, making it one of the fastest open ASR models available.

Architecture

Encoder-decoder architecture with a Conformer-based audio encoder and a Qwen3 0.6B decoder. The audio encoder processes mel-spectrogram features through conformer blocks with relative positional encoding, while the decoder generates text tokens autoregressively with cross-attention to encoder states.

Mixpeek SDK Integration

// No extractor parameter takes a Hugging Face model id (checked against
// GET /v1/discovery/extractors, which returns 13), so Qwen3-ASR-0.6B runs
// on your side and the output is upserted through POST
// /v1/namespaces/{namespace_id}/documents/upsert. On Enterprise the other
// path is to upload the weights instead: POST /v1/namespaces/{id}/models
// accepts the huggingface format and a custom plugin loads them.
const res = await fetch(
  "https://api.mixpeek.com/v1/namespaces/ns_your_namespace/documents/upsert",
  {
    method: "POST",
    headers: {
      Authorization: "Bearer API_KEY",
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      collection_id: "col_your_collection",
      documents: [
        {
          document_id: "asset-00412",
          // The model produces text, so it lands in payload. Give the
          // collection a text vector index and embed that text to make it
          // searchable rather than only filterable.
          payload: { extracted_text: modelOutput, source_key: "archive/2026/asset-00412" },
          vectors: { "text-embedding": embeddingOfModelOutput },
        },
      ],
    }),
  },
);

// Managed alternative, if this exact model is not the requirement:
// universal_extractor@v1 runs google/gemini-embedding-2
// (3072-d) over a bucket, with no inference of your own.