Developers

Run Wujud's tools from code or an AI assistant

The same tools as the dashboard, at the same credit prices.

REST API

Create an API key in the dashboard and call the tools from your own app or script.

How it works →

MCP server

Claude, ChatGPT and Cursor run Wujud tools for you.

Connect a client →

llms.txt

A plain summary of Wujud written for AI, so assistants answer questions about it correctly.

Ask AI →

REST API

Run the same tools from your own code

Every tool on Wujud is also a REST endpoint. Send a request with your API key, get a job back, and check on it until the result is ready. A run costs the same credits as it does in the dashboard, and an estimate endpoint tells you the cost first.

  1. 1 Create an API key in workspace Settings
  2. 2 Send a request, get a jobId back (202)
  3. 3 Check /ai/jobs/{jobId} until it's done

Which parts of the API a key can reach depends on your plan.

POST Illustrative request shape
curl https://wujud.app/api/v1/ai/ocr \
  -H "X-API-Key: $WUJUD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "sourceUrl": "https://example.com/receipt.jpg",
    "documentType": "RECEIPT"
  }'

MCP server

Let your AI assistant run the tools

Connect Claude, ChatGPT or Cursor by signing in to Wujud. Your assistant can check your credits and run the AI tools in any workspace you own.

  1. 1 Add the Wujud MCP server to Claude, ChatGPT or Cursor
  2. 2 Sign in to Wujud and pick your default workspace
  3. 3 Ask your assistant to run a tool, in any workspace you own
See how to connect
MCP Then run /mcp to sign in. In Claude and ChatGPT, add the same URL as a connector
claude mcp add --transport http wujud \
  https://mcp.wujud.app/mcp

Ask your AI about Wujud

Pick an assistant and it opens with the question already asked, pointed at our llms.txt, a plain summary of Wujud written for AI. Then ask whatever you want to know.