---
title: Runs | FLORA API
description: Start generation and technique runs, then poll for status and outputs.
---

Runs are asynchronous workflow executions. Create a run, keep the returned run ID or poll URL, then poll until the run reaches `completed` or `failed`.

## Create a technique run

The slug-based technique run endpoint is the most common path:

Terminal window

```
curl -X POST https://app.flora.ai/api/v1/techniques/cctv-cam/runs \
  -H "Authorization: Bearer sk_live_XXXX" \
  -H "Content-Type: application/json" \
  -d '{
    "inputs": [
      { "id": "input_image", "type": "imageUrl", "value": "https://example.com/photo.png" }
    ],
    "mode": "async",
    "idempotency_key": "request-123"
  }'
```

Response:

```
{
  "runId": "run_...",
  "status": "pending",
  "progress": 0,
  "pollUrl": "https://app.flora.ai/api/v1/techniques/cctv-cam/runs/run_..."
}
```

## Poll a technique run

Terminal window

```
curl https://app.flora.ai/api/v1/techniques/cctv-cam/runs/run_XXXX \
  -H "Authorization: Bearer sk_live_XXXX"
```

Poll every 2–5 seconds. Stop when `status` is `completed` or `failed`.

## Start a general generation run

Some workflows use the top-level generation endpoint:

Terminal window

```
curl -X POST https://app.flora.ai/api/v1/runs/generation \
  -H "Authorization: Bearer sk_live_XXXX" \
  -H "Content-Type: application/json" \
  -d '{
    "type": "image",
    "prompt": "A small blue flower on a white background",
    "workspace_id": "ws_XXXX",
    "project_id": "proj_XXXX"
  }'
```

The response includes a run identifier such as `run_id`, `runId`, or `id`, and may include a `workflow_run_id`.

## Start a top-level technique run

If you already have a technique ID from `GET /api/v1/techniques/{slug}`, you can also start a run through the top-level technique endpoint:

Terminal window

```
curl -X POST https://app.flora.ai/api/v1/runs/technique \
  -H "Authorization: Bearer sk_live_XXXX" \
  -H "Content-Type: application/json" \
  -d '{
    "technique_id": "tech_XXXX",
    "workspace_id": "ws_XXXX",
    "inputs": {
      "input_image": "https://example.com/photo.png"
    }
  }'
```

## Run statuses

| Status      | Meaning                               |
| ----------- | ------------------------------------- |
| `pending`   | The run is queued                     |
| `running`   | Work is in progress; check `progress` |
| `completed` | Outputs are ready                     |
| `failed`    | Check `errorCode` and `errorMessage`  |

## Common endpoints

| Method | Path                                     | Description                           |
| ------ | ---------------------------------------- | ------------------------------------- |
| POST   | `/api/v1/techniques/{slug}/runs`         | Create a slug-based technique run     |
| GET    | `/api/v1/techniques/{slug}/runs/{runId}` | Poll a slug-based technique run       |
| POST   | `/api/v1/runs/generation`                | Start a general generation run        |
| POST   | `/api/v1/runs/technique`                 | Start a technique run by technique ID |

## Notes

- Runs are asynchronous. Do not expect final outputs from the create response.
- Use `idempotency_key` for slug-based technique runs when retrying after network errors.
- Permission or credit constraints may return `403 forbidden` or `402 insufficient_credits`.
- Output URLs are long-lived but not permanent. Download anything you need to keep.