
Three.Ws Vision
- 89 repo stars
- Updated July 28, 2026
- nirholas/three.ws
Gives an AI agent image understanding: analyze and describe any image via the three.ws vision pipeline, with an anonymous free tier.
About
three.ws Vision is an MCP server that gives an AI agent the ability to look at images: it calls the three.ws vision pipeline to analyze and describe any image you pass it. A solo builder reaches for it when their Claude Code or Cursor agent needs to reason over screenshots, diagrams, or photos without wiring up a vision model directly. It runs over npx stdio, works anonymously on a free per-IP rate limit, and takes an optional API key to raise limits.
- Image analysis and description for agents
- Anonymous free per-IP tier, optional API key
- npx stdio server, configurable base URL
Three.Ws Vision by the numbers
- Data as of Jul 28, 2026 (Skillselion catalog sync)
claude mcp add --env THREE_WS_BASE=YOUR_THREE_WS_BASE --env THREE_WS_TIMEOUT_MS=YOUR_THREE_WS_TIMEOUT_MS --env THREE_WS_API_KEY=YOUR_THREE_WS_API_KEY vision-mcp -- npx -y @three-ws/vision-mcpAdd your badge
Show developers this MCP server is listed on Skillselion. Paste this into your README.
| repo stars | ★ 89 |
|---|---|
| Package | @three-ws/vision-mcp |
| Transport | STDIO |
| Auth | None |
| Last updated | July 28, 2026 |
| Repository | nirholas/three.ws ↗ |
What it does
Gives an AI agent image understanding: analyze and describe any image via the three.ws vision pipeline, with an anonymous free tier.
Who is it for?
Agents that need to analyze or caption images without hosting a vision model
Skip if: Offline or air-gapped setups (requires the hosted three.ws API)