Schema | Local Vision MCP

Local Vision MCP

Describes the environment variables required to run the server.

Name	Required	Description	Default
`OLLAMA_HOST`	No	Ollama service address	http://localhost:11434
`DETECTION_MODEL`	No	Default COCO detection model	yolov8n.pt
`MOBILECLIP_TS_URL`	No	URL to download mobileclip_blt.ts (defaults to a mirror)
`VISION_OUTPUT_DIR`	No	Set to force all generated files to be written to this directory
`SEGMENTATION_MODEL`	No	Default segmentation model	yolov8n-seg.pt
`VISION_MODEL_QUICK`	No	Quick mode model (falls back to 8B if not installed)	qwen3-vl:4b
`LOCAL_VISION_MAX_MB`	No	Maximum single image size in MB	20
`OLLAMA_VISION_MODEL`	No	Default vision model	qwen3-vl:8b
`DETECTION_TEXT_MODEL`	No	Default zero-shot detection model	yoloe-v8s-seg.pt
`PADDLEOCR_KEEP_ONEDNN`	No	Set to '1' to keep PaddleOCR's oneDNN (requires paddle version without bug or GPU version)

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Server capabilities have not been inspected yet.

Functions exposed to the LLM to take actions

Name	Description
No tools

Interactive templates invoked by user choice

Name	Description
No prompts

Contextual data attached and managed by the client

Name	Description
No resources

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Yuhang-uestc/deepvision-local-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server