Research
The GLM Large Model Family
All
Base Model
Mutimodal AI
Reasoning Model
Agentic LLM
Code Model
Time sorting
Base Model
2026/06/16
GLM-5.2: Built for Long-Horizon Tasks
Base Model
2026/05/20
Next-generation LLM Inference Network: How ZCube Alleviates Network Bottlenecks?
ZCube：Network really matters for LLM inference
Base Model
2026/04/29
Scaling Pain of Coding Agent Serving: Lessons from Debugging GLM-5 at Scale
we share the lessons learned from this investigation, with the hope of helping the community better understand and overcome the Scaling Pain of Coding Agent inference.
Base Model
2026/04/07
GLM-5.1: Towards Long-Horizon Tasks
our next-generation flagship model for agentic engineering
Mutimodal AI
2026/04/01
GLM-5V-Turbo: A Multimodal Coding Foundation Model
A multimodal coding foundation model for visual programming
Agentic LLM
2026/03/15
GLM-5-Turbo: A Foundation Model Enhanced for OpenClaw
A foundation model for OpenClaw
Base Model
2026/02/21
GLM-5 Technical Report
to transition the paradigm of vibe coding to agentic engineering
Base Model
2026/02/11
GLM-5: From Vibe Coding to Agentic Engineering
targeting complex systems engineering and long-horizon agentic tasks
Mutimodal AI
2026/02/02
GLM-OCR: SOTA Performance, Mastering Complex Document Recognition
a lightweight professional OCR model
Base Model
2026/01/19
GLM-4.7-Flash, open source and free
Small but powerful.
Mutimodal AI
2026/01/13
GLM-Image: Auto-regressive for Dense-knowledge and High-fidelity Image Generation
the first open-source, industrial-grade discrete auto-regressive image generation model
Mutimodal AI
2025/12/10
GLM-TTS: Controllable & Emotion-Expressive Zero-shot TTS with Multi-Reward Reinforcement Learning
Generating more expressive and emotional speech.
Mutimodal AI
2025/12/09
GLM-ASR-Nano: Robust Speech Recognition for the Real World
a robust, open-source speech recognition model with 1.5B parameters
Agentic LLM
2025/12/08
AutoGLM Goes Open Source
Unlocking the AI Phone for Everyone
Mutimodal AI
2025/12/07
GLM-4.6V: Open Source Multimodal Models with Native Tool Use
effectively bridges the gap between "visual perception" and "executable action"