Short answer: Drop the "Replicate → Run Replicate Prediction" action anywhere in your workflow, map the inputs from upstream nodes, and publish.
Every field can be mapped from an upstream trigger, AI step, table row, or hard-coded literal.
| Field | Type | Required | Description |
|---|---|---|---|
Model Version model_version | string | Required | Full model version string: owner/model:version_hash |
Input (JSON) input | string | Required | Model input parameters as JSON. Varies by model. |
Wait for Result wait | options | Optional | Wait synchronously for the prediction to complete (up to 60s) |
{"model_version": "e.g. stability-ai/sdxl:abc123...","input": "{\"prompt\": \"A photo of an astronaut riding a horse\"}","wait": "{{trigger.wait}}"}
{"id": "pred123","output": ["https://replicate.delivery/output123.png"],"status": "succeeded","metrics": {"predict_time": 3.5}}
Use these fields in downstream nodes for routing, logging, or error handling.
Any of these apps can fire this action as part of a workflow.
Triggered by anything in the catalog. Free tier available. No credit card.