GLM-5.2

GLM-5.2

Your AI assistant — ask anything, attach files, write code.

📊 Complex Analysis
Detailed tables with multi-column comparisons
💻 Code Generation
Syntax-highlighted blocks with explanations
📋 Structured Planning
Phased breakdowns with collapsible sections
📈 Data Analysis
Rich tables with trend analysis
🔌
Ollama is not running
Start Ollama to chat with GLM-5.2:
ollama run glm-5.2:cloud
GLM-5.2 can make mistakes. Verify important info.

Settings

Model

Available models are loaded from Ollama

Connection

Leave blank to use the bundled proxy (works when running python3 server.py locally). On the hosted demo, paste your Ollama URL to chat — Ollama needs CORS on: run OLLAMA_ORIGINS=* ollama serve.

System Prompt

Leave empty to use the default prompt. This applies to all new messages.

Preferences

Hosted demo — chat with your own Ollama from this page: open Settings → Ollama URL and paste http://localhost:11434 (run Ollama with OLLAMA_ORIGINS=*). View the repo →

Connect your Ollama to start chatting

No Ollama backend is reachable from this page yet. Point it at your own Ollama in two steps.

1. Run Ollama with browser access on
OLLAMA_ORIGINS=* ollama serve
This lets a web page call Ollama directly. Without it the browser blocks the request.
2. Enter your Ollama URL
On your own machine it's usually http://localhost:11434. On another device on your network, use that machine's IP.