GLM-5.2: Built for Long-Horizon Tasks

Hugging Face Blog

ZAI released GLM-5.2 as an open-weights model with significant improvements in coding and agentic tasks. It ranks #3 overall on the GDPval-AA agentic benchmark, ahead of GPT-5.5. Reports indicate it runs at 120 tok/s across two networked Blackwell tinyboxes.

GLMOpen source

See 14 more sources

zai-org/GLM-5.26 days agozai-org

unsloth/GLM-5.2-GGUF5 days agounsloth

zai-org/GLM-5.2-FP86 days agozai-org

GLM is the kind of model that revives serious interest in open source AI. It passes the blind test...13 hours agoAravind Srinivas

RT @Thom_Wolf: Desert island survival list: ✅ Solar panel / battery ✅ 256 GB Mac Studio ✅ GLM 5.2...2 days agoHugging Face

RT @Yuchenj_UW: After using GLM-5.2 for a day, I’m surprised by how often it feels close to Opus...3 days agoAravind Srinivas

GLM now supported on Perplexity Agent API4 hours agoAravind Srinivas

GLM-5.2 and IndexShare for Long-Context Sparse Attention4 days agoSebastian Raschka

GLM-5.2 is the step change for open agents18 hours agoNathan Lambert

[AINews] not much happened today3 days ago

[AINews] GLM-5.2: the top Frontend Coding model in the world, IndexShare for Speculative Decoding6 days ago

GLM-5.2 Is The New Best Open Model20 hours agoZvi Mowshowitz

GLM-5.2 is probably the most powerful text-only open weights LLM5 days agoSimon Willison

GLM-5.2 is the step change for open agents. Aka GLM-5.2 is the “DeepSeek moment” for agents. We ente...18 hours agoNathan Lambert

Hugging Face Blog

Jun 17, 9:06 AM