Qwen3-4B-Thinking-2507 (BaseRT)
BaseRT .base builds of
Qwen/Qwen3-4B-Thinking-2507 for fast local inference on Apple Silicon.
| File | Quant |
|---|---|
Qwen3-4B-Thinking-2507-Q4.base |
4-bit (default) |
Qwen3-4B-Thinking-2507-Q8.base |
8-bit |
basert serve basecompute/Qwen3-4B-Thinking-2507 # Q4 (default)
basert serve basecompute/Qwen3-4B-Thinking-2507 --variant default-q8 # Q8
Documentation: https://docs.basecompute.co