Tencent Hunyuan Releases T1, an RL-Trained Terminal Agent

Tencent Hunyuan released T1, a 122B-parameter MoE (10B active) model trained with reinforcement learning to operate real terminals on long-horizon tasks. It raised Terminal-Bench 2.1 pass rate from 43.8% to 64.0%.

2026-09-11 ~ 2026-09-11 · 2 related posts

• Tencent Hunyuan releases T1: 122B MoE model RL-trained for long-horizon terminal tasks — Tencent-Hunyuan · 2026-09-11
• T1: RL-trained 122B MoE terminal agent lifts Terminal-Bench 2.1 from 43.8% to 64.0% — heghbalz · 2026-09-11