Alibaba's Qwen3.6-35B-A3B hits 24.6% on Terminal-Bench 2.0 with just 3B active params

Alibaba's Qwen team released Qwen3.6-35B-A3B — a sparse MoE model with 35B total parameters but only 3B active per token — that hit 24.6% on the public Terminal-Bench 2.0 leaderboard. The Qwen team also released Qwen3.5-9B alongside it. The 35B/3B model is being highlighted as a budget-friendly choice for developers building AI products on tight inference economics.