Model Overview
| Model | Model ID | Parameters | Context | Features |
|---|---|---|---|---|
| GLM-4.6V | glm-4.6v |
- | 128K | Image + Video Understanding |
| GLM-4.1V-Thinking | glm-4.1v-thinking |
- | 128K | Visual Reasoning, Chain-of-Thought |
| GLM-OCR | glm-ocr |
0.9B | 8K | Dedicated OCR, OmniDocBench 94.62 SOTA |
| AutoGLM-Phone | autoglm-phone |
- | - | Phone Automation Control |
Selection Recommendations
- General Image Understanding:
glm-4.6v - Video Analysis:
glm-4.6v - Document OCR:
glm-ocr— 0.9B parameters, 94.62 SOTA - Visual Reasoning:
glm-4.1v-thinking— Chain-of-Thought - GUI Agent/Frontend:
glm-4.1v-thinking— Supports visual grounding, frontend code generation - Phone Automation:
autoglm-phone
API Key Configuration
From ~/.glm/.env load ZHIPU_API_KEY. Key acquisition:https://open.bigmodel.cn/usercenter/apikeys
GLM-OCR Features
- 0.9B parameters, OmniDocBench 94.62 SOTA
- Supported languages: Chinese, English, Japanese, Korean, etc.
- Output formats: Markdown, LaTeX, HTML
- Layout Parsing API Endpoint
AutoGLM-Phone Features
- Android Device Control
- Automated Operation Execution
- Dedicated API Endpoint
Recommended Scenarios
Image understanding, video analysis, document Q&A, GUI Agent, phone automation