Agent Scripting API 429s during high-concurrency load test

Context:
Running JMeter script to validate agent scripting API throughput. Thread group set to 300 concurrent users attempting to update scripts. Environment is Genesys Cloud Production. Seeing 429 Too Many Requests errors after 15 seconds. Rate limit headers indicate 1000 requests per minute window.

Question:
Why does this setting block valid concurrent updates from distinct agent IDs? Is there a global org limit or a per-user token cap I am missing?