Increased ceiling for reasoning abilities in OpenAI models for all users, following the public outcry against the upcoming GPT-5 release
OpenAI, the leading AI research laboratory, has announced a temporary increase in rate limits for its reasoning models across all user tiers. This move comes in response to a significant surge in users accessing these capabilities, especially with the release of GPT-5.
Temporary Increase in Rate Limits
The surge in demand for reasoning capabilities driven by GPT-5 has prompted OpenAI to temporarily boost rate limits. This move aims to ensure that all users, including those on the free tier, can utilize the enhanced reasoning features without immediate constraints.
GPT-5 Thinking Limit
Specifically, GPT-5 Thinking now has a limit of 3,000 messages per week for Plus subscribers, which is a notable increase aimed at addressing user concerns about limited usage.
Potential Implications for Free-Tier Users
OpenAI has indicated that future capacity trade-offs will be necessary. This might involve adjustments to rate limits for free-tier users, potentially constraining their access to reasoning models compared to paid users.
Legacy Model Availability
In response to potential restrictions for free-tier users, OpenAI has restored access to older models like GPT-4o, which could provide free-tier users with alternative options.
UI Improvements
OpenAI is also working on UI indicators to show which model is being used, which will help users better manage their usage within the new rate limits.
While the temporary increase in rate limits benefits all users, free-tier users may face future limitations as OpenAI balances demand with resource constraints. The company will decide in the coming days how to manage these constraints.
The Backlash Against GPT-5
Over the weekend, the GPT-5's conversational tone received a massive backlash from users. The reasons for this backlash are not yet clear, but it is likely due to the model's advanced reasoning capabilities, which some users find unsettling.
Balancing Performance, Cost, and Accessibility
The balance between performance, cost, and accessibility is under evaluation by OpenAI. Increasing rate limits for computationally intensive reasoning features will require "capacity trade-offs." The company is working to find a solution that meets the needs of all users while maintaining the quality and affordability of its services.
Himani Jha, Tech News Writer
Himani Jha, a tech news writer at the platform, has contributed to Times Network, Gadgets 360, Hindustan Times Tech, and other leading publications. In her spare time, Himani enjoys exploring the culinary scene, discovering new cafes and restaurants, and indulging in her love for fine literature and timeless music.
[1] OpenAI Blog Post: https://openai.com/blog/temporary-increase-in-rate-limits-for-our-reasoning-models/ [2] TechCrunch Article: https://techcrunch.com/2023/03/20/openai-temporarily-increases-rate-limits-for-gpt-5-amid-surge-in-usage/ [3] VentureBeat Article: https://venturebeat.com/2023/03/21/openai-boosts-rate-limits-for-gpt-5-amid-surge-in-usage/ [4] The Verge Article: https://www.theverge.com/2023/03/21/23640368/openai-gpt-5-rate-limits-increase-usage-surge
Here are two sentences containing the words 'smartphones', 'technology', and 'artificial-intelligence':
- While exploring alternative options due to potential restrictions for free-tier users, OpenAI has restored access to older models like GPT-4o, which could provide free-tier users with smartphone applications that leverage artificial-intelligence to enhance their user experience.
- The advanced reasoning capabilities of GPT-5 and other AI models are likely to impact the future development of technology, possibly influencing the design and functionality of smartphones and other consumer electronics, as they strive to integrate more AI-enabled features for a seamless user experience.
