Skip to main content

Qwen3 Embedding 8B

High-quality multilingual text embeddings (8B).

Overview​

  • model_name: hf-Qwen-Qwen3-Embedding-8B (HuggingFace import, currently free)
  • Source: Qwen/Qwen3-Embedding-8B
  • Task: text embeddings — top-tier retrieval quality (MTEB leaderboard family)
  • Served on: GPU — verified on Intel Arc (XPU) nodes
  • Status: ready (verified end-to-end through the public inference route, 2026-08-02)

When to use​

✅ Semantic search, RAG retrieval, clustering, and deduplication where embedding quality matters more than latency. Pairs naturally with Qwen3 Reranker 8B for a retrieve-then-rerank pipeline.

❌ For high-throughput / low-cost embedding of large corpora, the lighter specialist-embeddings endpoint is cheaper per token.

Input contract​

{"input": {"texts": ["first text", "second text"]}}

A single text string is also accepted and treated as a one-element batch. The response contains one embedding vector per input text.

Quick start​

curl -X POST "https://api.colabhive.com/api/builder/v1/endpoints/<endpoint_id>/infer" \
  -H "X-API-Key: $COLABHIVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"input": {"texts": ["hello world", "embedding quality test"]}}'