Qwen3.6-35B-A3B and 9B land on the public Terminal-Bench 2.0 leaderboard
Qwen3.6-35B-A3B and 9B are officially on the public Terminal-Bench 2.0 leaderboard!
little-coder × Qwen3.6-35B-A3B scored 24.6% ±3.2 on Terminal-Bench 2.0, above Gemini 2.5 Pro on Gemini CLI at 19.6% and Qwen3-Coder-480B on Terminus 2 at 23.9%.
Why it matters: Qwen3.6 in 35B-A3B and 9B sizes lands on the public Terminal-Bench 2.0 leaderboard, giving small models a checkable public score on hard terminal-agent tests.