Qwen3-Reranker-0.6B

Name: Qwen3-Reranker-0.6B
Author: Qwen

by Qwen

Compact 0.6B text reranker with 100+ language support

1.4Mdl/month

0.6Bparams

HuggingFace Run it on your own data, free

Identifiers

Model ID

Qwen/Qwen3-Reranker-0.6B

Feature URI

mixpeek://reranker@v1/qwen3_reranker_06b_v1

Overview

Qwen3-Reranker-0.6B is the smallest model in the Qwen3 reranker family, supporting over 100 languages with instruction-following capabilities. Despite its compact size, it delivers competitive reranking quality through knowledge distillation from larger Qwen3 models. Its small footprint makes it ideal for high-throughput reranking where latency matters.

Architecture

Decoder-based cross-encoder built on the Qwen3 0.6B language model. Uses instruction tuning to accept natural-language reranking instructions, allowing users to specify custom relevance criteria. Produces relevance scores by encoding query-document pairs with a classification head.

Mixpeek SDK Integration

// Reranking is a retriever STAGE in Mixpeek, not an ingest-time extractor.
// The rerank stage runs a cross-encoder inference service; the shipped default
// is BAAI/bge-reranker-v2-m3. Pointing it at Qwen3-Reranker-0.6B means registering that
// model as a custom reranker plugin and naming it in feature_uri, which is an
// Enterprise path. Stage contract read from GET /v1/discovery/stages.
const retriever = await mx.retrievers.create({
  namespace_id: "my-namespace",
  retriever_name: "search-then-rerank",
  stages: [
    {
      stage_name: "candidates",
      stage_id: "feature_search",
      parameters: { limit: 100 },
    },
    {
      stage_name: "rerank_results",
      stage_id: "rerank",
      parameters: {
        inference_name: "BAAI__bge_reranker_v2_m3",
        query: "{{INPUT.query}}",
        document_field: "content",
        top_k: 10,
      },
    },
  ],
});