Copilot CLI Rate Limits: Hindering Agentic Coding and Developer Performance
Unpacking the Frustration: Copilot CLI's Unexpected Rate Limits
A recent discussion on GitHub's community forums has shed light on a significant hurdle for users adopting GitHub Copilot Pro, specifically when interacting via the Copilot CLI. Developers are encountering frequent and restrictive rate limits, severely impacting their ability to leverage this promising performance development tool for agentic coding.
The issue, brought to the forefront by user galluccipablo, details an experience where Copilot CLI became virtually unusable due to persistent API errors and rate limiting messages, often within moments of starting a new session.
The Core Problem: Hitting Limits Too Soon
The original poster, a new Copilot Pro subscriber, reported hitting rate limits on their second day of use, and as early as their second request. This rapid exhaustion of API requests manifests primarily as a recurring message: Request failed due to a transient API error. Retrying... While this message suggests a temporary glitch, logs reveal a more specific and concerning underlying cause: HTTP 429 status codes with the message user_model_rate_limited.
Only once did the explicit rate limit message surface directly in the UI:
Sorry, you've hit a rate limit that restricts the number of Copilot model requests you can make within a specific time period. Please try again in 1 minute. Please review our Terms of Service (https://docs.github.com/site-policy/github-terms/github-terms-of-service). (Request ID: 9666:1776BE:2067F6:2C5E97:69C85B2B)This behavior is particularly frustrating for users who describe their usage as "very light" compared to other AI coding assistants like Claude or Codex. The discrepancy raises questions about the current rate limit policies and their practical implications for an efficient engineering workflow.
Impact on Engineering Workflow and Developer Software Evaluation
For developers evaluating Copilot Pro as a replacement for other coding plan subscriptions, these stringent rate limits render agentic coding "unusable." The promise of an AI assistant that can autonomously perform tasks and generate code is severely undermined if the underlying developer software cannot sustain even moderate usage. This directly impacts developer productivity and the perceived value of Copilot Pro as a robust performance development tool.
GitHub's Acknowledgment (But No Immediate Fix)
The discussion received an automated response from github-actions, acknowledging the feedback and assuring that it would be reviewed by product teams. While this confirms that the feedback has been submitted, it does not offer an immediate solution, workaround, or timeline for addressing the rate limit issue. Users are encouraged to monitor the Changelog and Product Roadmap for updates, and to continue engaging with the community.
Community Call: Share Your Experience
This ongoing issue highlights a critical area for improvement for GitHub Copilot CLI. If you've experienced similar rate limiting issues with Copilot CLI or have found workarounds, your insights are invaluable. Share your experiences and contribute to the discussion at the original GitHub thread: Discussion #190996: Copilot CLI - Request failed due to a transient API error - rate limit. Your collective feedback is essential in shaping the future of this important developer tool.
