High-throughput Qwen3 Embedding 0.6B runtime with prepacked Q4 weights and custom WebGPU kernels
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).