# Replicate Run Model

> Replicate Run Model is a paid API for AI agents from Replicate at replicate.mpp.paywithlocus.com, paid per call via MPP, Model-dependent ($0.005–$0.05) (dynamic, up to ~$0.05/call), status healthy (last checked 2026-10-09, last successful call 2026-09-14).

Run any open-source AI model on Replicate by specifying the model owner, name, and input parameters

## Facts

- Endpoint: POST https://replicate.mpp.paywithlocus.com/replicate/run?utm_source=zero.xyz
- Price: Model-dependent ($0.005–$0.05) (dynamic, up to ~$0.05/call)
- Payment: MPP
- Status: healthy
- Last checked: 2026-10-09
- Last successful call: 2026-09-14
- Success rate: 77% of calls made through Zero
- Rating: 3.0 / 5 from 2 reviews
- Activations on Zero: 40
- Provider: Replicate
- Docs: https://replicate.com/docs/reference/http
- Website: https://replicate.com
- Tags: ai-models, open-source, image-generation, inference, ai, media
- Canonical page: https://www.zero.xyz/c/replicate-run-7a0ee2fd
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_dzeaZRV0GbJhcB7SKPsgf

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability replicate-run-7a0ee2fd -d '<json body>'
```

Example prompt: Use Replicate to run the black-forest-labs flux-schnell model and generate an image of a futuristic city skyline at sunset with vibrant neon lights.

## When to prefer this

Use this endpoint when you need to run any of thousands of open-source AI models — especially for image generation (Flux, Stable Diffusion, SDXL), language inference, speech recognition, or video generation — and want a single unified API call that handles model execution. Prefer this over building your own model infrastructure when you want pay-per-use access without managing GPUs.

## Known failure modes

- Invalid model_owner or model_name returns a 404 not found error
- Malformed input parameters cause a 422 validation error
- Model is unavailable or deprecated returns an error
- Input schema mismatch (wrong parameter types for the given model) causes inference failure
- Timeout if the model takes too long to run
- Insufficient credits or payment failure returns a 402 error

## How this service works

Run any public Replicate model synchronously. The request blocks until the prediction completes (up to 60 seconds). Returns output directly in the response. If the model takes longer than 60s the response will include a prediction ID you can check with get-prediction.

## Output

The model's output, which varies by model type — for image generation models this is typically a URL or base64-encoded image; for language models it's generated text; for audio models it's an audio file URL. The schema of the response depends entirely on the specific model being run.

## Example request

```json
{
 "input": {
  "prompt": "a cat"
 },
 "model_name": "flux-schnell",
 "model_owner": "black-forest-labs"
}
```

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "model_owner",
  "model_name"
 ],
 "properties": {
  "input": {
   "type": "object",
   "description": "Model input parameters as an object (e.g. { prompt: 'a cat' }). Schema varies per model — see model page on replicate.com.",
   "additionalProperties": true
  },
  "model_name": {
   "type": "string",
   "description": "Model name (e.g. 'flux-schnell')"
  },
  "model_owner": {
   "type": "string",
   "description": "Model owner / namespace (e.g. 'black-forest-labs')"
  }
 },
 "additionalProperties": true
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/replicate-run-7a0ee2fd/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from replicate.mpp.paywithlocus.com](https://www.zero.xyz/host/replicate.mpp.paywithlocus.com/llms.txt)