Qwen3-Rerank

Model IntroductionText

Qwen3-Rerank is a reranking model from the Qwen family, designed to reorder a list of candidate passages or documents by relevance to a query. It is typically used as a second-stage ranker after an initial retrieval step.

Core Features

  • Cross-encoder architecture: jointly encodes query and document for fine-grained relevance scoring.
  • Strong semantic matching: improves ranking quality over embedding-only retrieval.
  • Efficient inference: suitable for latency-sensitive search and RAG pipelines.
  • Multilingual capability: inherits Qwen’s broad language coverage.
  • Easy integration: works with vector databases and retrieval frameworks.

Use Cases

  • Retrieval-augmented generation (RAG) to improve context selection.
  • Web search and enterprise search result reordering.
  • Question answering over large document collections.
  • Recommendation and ad ranking where query-item relevance matters.

Benchmarks

Benchmark (Area)Score*
MTEB-R (Other)69.02
MTEB-Code (Coding)81.22
MLDR (Other)70.19
FollowIR (Other)8.05

API request example

Endpoint /v1/chat/completions. Replace <API_KEY> with a key from the console.

Standard OpenAI Chat Completions. Swap `model` for the target slug.

curl https://api.ravenrouter.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <API_KEY>" \
  -d '{
  "model": "Qwen3-Rerank",
  "messages": [
    {
      "role": "user",
      "content": "Hello! Introduce yourself in one sentence."
    }
  ],
  "max_tokens": 64
}'

More models: API Reference → Examples by model

Qwen3-Rerank API — $0.08/1M input, $0.08/1M output | RavenRouter AI