OpenAI Codex Rate Limit Reductions and Alternatives
AICodeKinggo watch the original →
the gist
OpenAI has silently halved Codex usage limits, likely to reallocate compute to enterprise partnerships. Developers are shifting to more stable alternatives like the GLM Coding Plan.
The Shift in OpenAI Compute Allocation
OpenAI has reduced Codex usage limits by 50 percent, a change that appears permanent given the company's silence over a 12-hour period following user reports. This reduction is likely driven by the launch of the OpenAI Deployment Company, a partnership initiative that requires dedicated compute resources for enterprise clients. Codex was primarily an acquisition tool for onboarding individual users, and the current strategy prioritizes high-value enterprise deployments over maintaining high-volume, loss-making individual access.
Recommended Alternatives for Developers
The GLM Coding Plan is currently the most stable alternative, maintaining consistent rate limits since its inception. It is compatible with various development tools, including Cline, OpenCode, and Crush. For users seeking other options, the following services are available:
- GLM Coding Plan: Offers a yearly subscription for $345, which provides long-term stability and migration support for users transitioning from other services.
- Verdant: Features a manager option that handles multi-task orchestration within a single thread.
- CommandCode: Provides low-cost entry points, including a $1 plan that includes access to DeepSeek V4 Pro.
- Kimi Coding Plan: Offers tiered pricing ranging from $19 to $199 per month.
- Antigravity Plans: Includes a $20 monthly plan, though it requires using the proprietary Antigravity editor.