Skip to main content
Baseten enforces per-API-key rate limits on the management API. The limits protect shared infrastructure, such as the build queue, from bursty automation.

​
Limits

The default limit applies to every /v1/* endpoint. A few endpoints that touch shared build and deployment infrastructure have stricter limits.
EndpointLimit
POST /v1/models/{model_id}/deployments/development/activate20 requests/minute
POST /v1/models/{model_id}/deployments/production/activate20 requests/minute
POST /v1/models/{model_id}/deployments/{deployment_id}/activate20 requests/minute
POST /v1/models/{model_id}/deployments/development/deactivate20 requests/minute
POST /v1/models/{model_id}/deployments/production/deactivate20 requests/minute
POST /v1/models/{model_id}/deployments/{deployment_id}/deactivate20 requests/minute
POST /v1/models/{model_id}/deployments/development/retry10 requests/minute
POST /v1/models/{model_id}/deployments/production/retry10 requests/minute
POST /v1/models/{model_id}/deployments/{deployment_id}/retry10 requests/minute
All other /v1/* endpoints100 requests/second
Baseten tracks each endpoint separately.

​
Rate-limited responses

A request over the limit returns 429 Too Many Requests: 
{
  "error": "Rate limit exceeded. Please try again later.",
  "retry_after": 37
}
retry_after is the number of seconds until the current rate-limit window resets. Wait at least that long before retrying.

​
Retry handling

For CI pipelines or scripts that call the management API in a loop, handle 429 explicitly: 
import time
import requests

def post_with_retry(url, headers, max_attempts=5):
    for _ in range(max_attempts):
        response = requests.post(url, headers=headers)
        if response.status_code != 429:
            return response
        retry_after = response.json().get("retry_after", 1)
        time.sleep(retry_after)
    return response
Back off on retry_after instead of retrying immediately. A tight retry loop wastes API calls; the server rejects every request until the window resets.

​
Request higher limits

If your workload needs sustained throughput above the default limits, contact support to request per-endpoint increases for your organization.