Qwen3.8 27B Benchmark: Artificial Analysis Says 52 Then, 34 Now

Qwen3.8 27B Benchmark: Artificial Analysis Says 52 Then, 34 Now

Qwen3.8 27B really did score 52 on the Artificial Analysis Intelligence Index — that was the correct figure on 2026-08-18, under index version v4.1.1, and it topped that board at #1 of 135. It also scores 34 today under v4.3.2. The model never changed; the index did. That two-sentence answer is worth more than the number itself, because almost every article ranking for this keyword still prints “52” from an August snapshot and has never re-checked the live page. This post walks through what the Qwen3.8 27B benchmark actually says on 2026-10-01, why the total moved, and — more usefully — why the shape of the profile never moved at all. ...

October 1, 2026 · 14 min · baeseokjae
Qwen 3.6-35B-A3B Local Deployment Guide: How 3B Active Params Changes the Build

Qwen 3.6-35B-A3B Local Deployment Guide: How 3B Active Params Changes the Build

If you want a frontier coding model without cloud bills, Qwen3.6-35B-A3B is the first local LLM I’d run right now because its 3B active-parameter MoE path behaves like a 3B model in many deployment budgets while preserving much stronger coding signal than smaller dense models. I am going to show you the exact setup flow I use for local inference: runtime choice, quantization, context strategy, and verification commands that keep inference stable. ...

June 11, 2026 · 13 min · baeseokjae