Use this endpoint to call the development deployment of your model asynchronously.
Api-Key
(e.g. {"Authorization": "Api-Key abcd1234.abcd1234"}
)./async_predict
request payloads.
webhook_endpoint
is empty, your model must save prediction outputs so they can be accessed later. priority
is between 0 and 2, inclusive.max_time_in_queue_seconds
must be between 10 seconds and 72 hours, inclusive./async_predict
endpoint are limited to 200 requests per second.
QUEUED
or IN_PROGRESS
async requests, summed across all deployments.
/async_predict
requests will receive a 429 status code.
To avoid hitting these rate limits, we advise:
/async_predict
with exponential backoff in response to 429 errors.