Model · Alibaba · Reranking
Qwen3-Reranker (0.6B / 4B / 8B)
Alibaba's open rerankers under Apache 2.0 — the second-pass model that re-scores search results so the best matches rise to the top. Three sizes (0.6B, 4B, 8B), 100+ languages, 32K context, and task instructions you can customise.
- Modality
- Reranking
- License
- Apache 2.0 (Open)
- Parameter size
- 0.6B, 4B and 8B (4B is the most downloaded)
- Context window
- 32,768 tokens 32K tokens on all three sizes.
- Released
- June 5, 2025
- Last verified
- October 28, 2026
- Runs locally
- Yes
What you need to run it
- Mac
- Yes. Any Apple Silicon Mac (M1 or newer) with at least 8 GB (16 GB recommended) of memory. Newer chips are faster, but memory is what decides whether it runs.
- Windows / Linux PC
- NVIDIA works best. A graphics card with at least 4 GB (12 GB recommended) of video memory.
- No graphics card
- Works. Runs on a normal computer's processor, just slower.
The 0.6B size runs on a laptop; the 4B is the better-quality pick for a 16 GB Mac or a 12 GB graphics card.
Numbers are for: 0.6B at 16-bit for the minimum; 4B for the recommended figures (MLX builds exist for 0.6B, 4B, 8B). Our estimate from the model’s size — the maker publishes no figure.
Strengths
- Apache 2.0 — no commercial restrictions
- Instruction-aware: tell it what the search is for and results improve by about 1-5%, per Alibaba
- 100+ languages plus code retrieval
- Built to pair with Qwen3-Embedding for a full self-hosted search pipeline
- The 0.6B runs on modest hardware; 4B and 8B trade speed for accuracy
Weaknesses
- Reranking is always a second pass — you still need an embedding model or keyword search first
- Text only
- 4B and 8B need a graphics card for responsive search
Try it
| Where | Type | Notes |
|---|---|---|
| Hugging Face (0.6B) | weights | Apache 2.0 |
| Hugging Face (4B) | weights | Apache 2.0; 8B also published |
| Sentence Transformers | local | CrossEncoder support |
Official sources
- Model card (0.6B) model
- GitHub github
- Qwen blog announcement
- Paper paper
Change log
- — Added hardware requirements (Mac support, memory, NVIDIA). Source: estimate from size and community builds (HF Qwen/Qwen3-Reranker-4B).
- — Initial entry. Added 2026-10-28 from the popularity sweep: 1.50M Hugging Face downloads/month for Qwen3-Reranker-0.6B and 2.25M for 4B. Released 2025-06-05 with Qwen3-Embedding, per the Qwen blog and the arXiv paper. Apache 2.0, 32K context, free to self-host.
Esc