Qwen3 Embedding 0.6B

Qwen3 Embedding 0.6B is the ultra-compact embedding model in the Qwen3 line — only 600M parameters, optimised for high-throughput RAG over a 32K context window. Compared to BGE-M3, it trades broader multilingual coverage for higher Arabic-specific quality and a much larger context. Use as the primary embedder when GCC/Arabic content dominates your corpus.

Quick start

Use any OpenAI-compatible client. Just change base_url and your key.

from openai import OpenAI

client = OpenAI(
    api_key="tl-xxxxxxxxxxxxxxxxxxxxxxxx",
    base_url="https://api.thalam.ai/v1",
)

response = client.chat.completions.create(
    model="qwen/qwen3-embedding-0.6b",
    messages=[
        {"role": "user", "content": "Hello from Thalam!"}
    ],
)

print(response.choices[0].message.content)