AnalysisAI ModelsSeptember 12, 2026

User compares Qwen-Next and 3.8 27B for coding on M5 Max

Reddit user running both models locally on an M5 Max with 128GB RAM reports Qwen-Next feels worse than 3.8 27B for coding in a Node.js pi agent. Both run via MTPLX, which the user says only offers a speed-optimized dynamic q4 with 8-bit attention.

1 source

More stories today

Open the live feed
User compares Qwen-Next and 3.8 27B for coding on M5 Max