Providers

The Providers API returns pre-built polling configurations for popular AI APIs. Use a provider’s id as the provider field when creating a task to automatically configure polling.

List providers

GET /v1/providers

No authentication required.

Example request

curl https://api.asyncqueue.io/v1/providers

Example response

{
  "providers": [
    {
      "id": "fal",
      "name": "fal.ai",
      "description": "AI media generation (images, video, audio)",
      "pollConfig": {
        "urlTemplate": "{startResult.body.status_url}",
        "interval": 3,
        "statusPath": "status",
        "doneValues": ["COMPLETED"],
        "failValues": ["FAILED"],
        "resultPath": "",
        "resultUrlTemplate": "{startResult.body.response_url}",
        "forwardAuth": true,
        "maxDuration": 600
      }
    },
    {
      "id": "replicate",
      "name": "Replicate",
      "description": "Run open-source AI models",
      "pollConfig": {
        "urlTemplate": "https://api.replicate.com/v1/predictions/{startResult.body.id}",
        "interval": 5,
        "statusPath": "status",
        "doneValues": ["succeeded"],
        "failValues": ["failed", "canceled"],
        "resultPath": "output",
        "forwardAuth": true,
        "maxDuration": 600
      }
    },
    {
      "id": "runwayml",
      "name": "RunwayML",
      "description": "AI video generation (Gen-3, Gen-4)",
      "pollConfig": {
        "urlTemplate": "https://api.dev.runwayml.com/v1/tasks/{startResult.body.id}",
        "interval": 5,
        "statusPath": "status",
        "doneValues": ["SUCCEEDED"],
        "failValues": ["FAILED", "CANCELLED"],
        "resultPath": "output",
        "errorPath": "failure",
        "forwardAuth": true,
        "maxDuration": 600
      }
    },
    {
      "id": "luma",
      "name": "Luma AI",
      "description": "Dream Machine video generation",
      "pollConfig": {
        "urlTemplate": "https://api.lumalabs.ai/dream-machine/v1/generations/{startResult.body.id}",
        "interval": 5,
        "statusPath": "state",
        "doneValues": ["completed"],
        "failValues": ["failed"],
        "resultPath": "assets.video",
        "errorPath": "failure_reason",
        "forwardAuth": true,
        "maxDuration": 600
      }
    },
    {
      "id": "elevenlabs-dubbing",
      "name": "ElevenLabs Dubbing",
      "description": "AI video/audio dubbing and translation",
      "pollConfig": {
        "urlTemplate": "https://api.elevenlabs.io/v1/dubbing/{startResult.body.dubbing_id}",
        "interval": 10,
        "statusPath": "status",
        "doneValues": ["dubbed"],
        "failValues": ["failed"],
        "resultPath": "dubbing_id",
        "errorPath": "error",
        "forwardAuth": true,
        "maxDuration": 1800
      }
    },
    {
      "id": "stability-video",
      "name": "Stability AI Video",
      "description": "Stable Video Diffusion image-to-video",
      "pollConfig": {
        "urlTemplate": "https://api.stability.ai/v2beta/image-to-video/result/{startResult.body.id}",
        "interval": 10,
        "statusPath": "_httpStatus",
        "doneValues": ["200"],
        "failValues": ["400", "404", "500"],
        "resultPath": "_binaryUrl",
        "errorPath": "errors",
        "forwardAuth": true,
        "maxDuration": 600
      }
    }
  ]
}

pollConfig fields

Field	Type	Description
urlTemplate	string	URL to poll, with {startResult.body.<path>} template variables
interval	number	Seconds between poll requests
statusPath	string	Dot-notation path to the status field in poll responses
doneValues	string[]	Status values that indicate the job completed successfully
failValues	string[]	Status values that indicate the job failed
resultPath	string	Dot-notation path to the result. Empty string returns the full response.
errorPath	string	Dot-notation path to error messages in failure responses
resultUrlTemplate	string	Optional separate URL to fetch the final result
forwardAuth	boolean	Whether to forward Authorization/X-API-Key headers to poll requests
maxDuration	number	Maximum seconds to poll before timing out