321 1 week ago

Uses the Q8_0 quantized version of the official Qwen/Qwen3.5 model files, without any other modifications.

vision tools thinking 0.8b 2b 4b 9b 27b