China's GLM-5.2 Drops as Open-Weight, MIT-Licensed — and Beats GPT-5.5 on SWE-bench Pro
Zhipu AI's GLM-5.2 scores 62.1% on SWE-bench Pro, surpassing GPT-5.5's 58.6%, ships with a 1M-token context window, and costs roughly a tenth of closed competitors. It's MIT-licensed and available for local deployment.
Zhipu AI released GLM-5.2 this week, and the numbers are hard to dismiss. As @BrianRoemmele highlighted, the model scores 62.1% on SWE-bench Pro — the increasingly standard benchmark for agentic coding — beating GPT-5.5's reported 58.6%. The model ships with MIT-licensed open weights, a 1M-token context window, and pricing that undercuts closed competitors by roughly 10x, according to @RoundtableSpace.
Unlock the full briefing
Get every story in today's briefing, the full archive, and the daily AI intelligence brief.
All stories today
Full archive
Daily brief
Cancel anytime. Payments powered by Stripe.