GLM-5.3

User reports Terminal Bench v4 scores for several models, with GLM-5.3 leading at 41.9%. Other results are GLM-5.3-Flash at 32.8%, DSV4.1-Flash at 26.8%, Qwen3.8-Flash-Next at 25.3%, DSV4-Pro at 14.1%, Kimi-K3 at 12.6%, DSV4-Flash at 12.1%, Qwen3.8-27B at 5.6%, Muse Glimmer at 0.5% and gemma4-31b at 0.0%. The user describes GLM-5.3 as in a league of its own among open models. No hardware or inference engine details are given.