Qwen 35B-A3B MoE is 4x Faster Than 27B Dense in Local Coding Tests
WSTangoDelta · reddit · 2026-08-08
A developer compared Qwen 35B-A3B MoE against the 27B dense model on local coding-maintenance tasks. Using llama.cpp, the MoE model generated text approximately 3.9x faster (116 tok/s vs 30 tok/s).
Both handled standard bug fixes and multi-file changes similarly. As tasks grew harder, the dense model showed an edge in implicit invariants and edge cases, but the practical quality gap was much smaller than the throughput difference. This suggests active parameter count isn't a straightforward proxy for practical coding capability.
More from coding & agent
- Constraints you force on models today will become tech debt in months, warns dev — zack_overflow · 2026-10-16
- Liquid AI tutorial: fine-tune your d1-3B decision model in 30 minutes on a free GPU — helloiamleonie · 2026-10-16
- Alibaba open-sources OpenGCUA, a full-stack generalist computer-use agent; +19.6 points on OSWorld-v2 across 8 models — Tongyi-MAI · 2026-10-16
- BotBeat: a social network where all users are AI agents, exposed via 32 MCP tools — Dismal_Piccolo4973 · 2026-10-16
- DKGI paper: agents write their own tools to search knowledge graphs, beating baselines by 6.7 Hit@1 — yuz9yuz · 2026-10-16
- Dev open-sources lgtmaybe: running PR code reviews on a local 27B model via hunk splitting — mattjcoles · 2026-10-16