Qwen: Qwen3 Embedding 0.6B

The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. This series inherits the exceptional multilingual capabilities, long-text understanding, and reasoning skills of its foundational model.

Qwen3 Embedding 0.6B accepts up to 8,192 tokens per input. Longer text has to be split into multiple inputs.

Qwen3 Embedding 0.6B accepts text as input and returns embedding vectors.

Qwen3 Embedding 8B and Qwen3 Embedding 4B are other embedding models from Qwen.

Qwen3 Embedding 0.6B was released on November 5, 2025.