ThinkingCap-Qwen3.6-27B: reduces overthinking, achieves 44 t/s on M5 Max

BottleCapAI released ThinkingCap-Qwen3.6-27B, benchmarked on M5 Max with fast speed (44 t/s) and fewer tokens, ideal for efficient inference.

• benchmark
• GGUF
• Qwen 3.6
• M5 Max
• ThinkingCap