create_prediction

Pack: replicate · Endpoint: https://gateway.pipeworx.io/replicate/mcp

Start a Replicate model run by submitting a version id and input object; returns a prediction id and status that can be polled with the prediction tool until output is ready.

Parameters

NameTypeRequiredDescription
versionstringyes
inputunknownyes
webhookstringno
webhook_events_filterarrayno
itemsstringno
streambooleanno

Example call

Arguments

{
  "version": "f178fa7a1ae43a9d672b0d1c6c1c0c5e1234abcd",
  "input": {
    "prompt": "a dog wearing sunglasses"
  }
}

curl

curl -X POST https://gateway.pipeworx.io/replicate/mcp \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"create_prediction","arguments":{"version":"f178fa7a1ae43a9d672b0d1c6c1c0c5e1234abcd","input":{"prompt":"a dog wearing sunglasses"}}}}'

TypeScript (@pipeworx/sdk)

import { Pipeworx } from '@pipeworx/sdk';
const pipeworx = new Pipeworx();

const result = await pipeworx.call('create_prediction', {
  "version": "f178fa7a1ae43a9d672b0d1c6c1c0c5e1234abcd",
  "input": {
    "prompt": "a dog wearing sunglasses"
  }
});

More examples

{
  "version": "f178fa7a1ae43a9d672b0d1c6c1c0c5e1234abcd",
  "input": {
    "prompt": "a dog wearing sunglasses"
  },
  "webhook": "https://example.com/webhook",
  "webhook_events_filter": [
    "completed"
  ],
  "stream": false
}

Connect

Add this to your MCP client config, or use one-click install buttons:

{
  "mcpServers": {
    "replicate": {
      "url": "https://gateway.pipeworx.io/replicate/mcp"
    }
  }
}

See Getting Started for client-specific install steps.

Regenerated from source · build August 28, 2026