GLM-4
ethanweave/glm4v-vision-mcp · model-adapters
GLM-4.6V Image Understanding MCP: Diagram/OCR/ Dialysis, Native Access to DeepSeek Harness (dsh-mcp-client), also compatible with Codex/Cline, etc.
★ 2GitHub stars
0Forks
2026-08-14Last updated
PythonLanguage
—License
Key features
- anallyze image (image, prompt, ...): text / image understanding, support URL, data URI, naked base64, local path, support up to 8 multi-graphs
- Add in /.dsh/profiles/ /cordis.pact. yml (note:command with your vv python absolute path)
- Achieving higher priority than normal dialogue
- 2 Confirm that fix models json.py has been executed (DeepSeek custom model directory hides MCP tool, issue #36382)
- ocr image (image, language, ...)OCR text recognition: line by line, save layout, table to markdown, support area cropping
Install command
dsh plugin --profile web add github:ethanweave/glm4v-vision-mcp