OpenAI has announced a significant performance overhaul for ChatGPT, with loading speeds increasing by 94% and the introduction of GPT-5.6 Multi-Agent v2, which automates task delegation among different models. These updates aim to eliminate user wait times and model selection hassles, enhancing the interactive experience.
TL;DR
- ChatGPT's loading speed has improved by 94%, with memory usage and network requests significantly reduced.
- GPT-5.6 Multi-Agent v2 allows automatic task delegation among different models, including Sol, Terra, and Luna.
- These updates position ChatGPT as a more efficient workflow platform, reducing the need for manual model selection and lowering reasoning costs.
What happened
OpenAI has rolled out a major performance update for ChatGPT, with loading speeds increasing by 94% and memory usage dropping by 41.2%. The number of network requests has plummeted by 98.2%, and conversation history loading time has been reduced by 99.6%.
The update was tested on an extreme conversation scenario with 741 rounds of interaction and a size of 231MB. The average opening time for such conversations has dropped from 27.62 seconds to 1.66 seconds, with memory growth reduced from 1030.7 MiB to 606 MiB.
OpenAI has also launched GPT-5.6 Multi-Agent v2, which allows the main Agent to automatically delegate subtasks to different models, including Sol, Terra, and Luna. Each sub-Agent can be independently set with its own reasoning intensity.
Why it matters
These updates make ChatGPT more responsive and efficient, particularly for developers running heavy Codex sessions. The interface is significantly more responsive, and memory usage has decreased, enhancing the overall user experience.
The introduction of GPT-5.6 Multi-Agent v2 reduces the need for manual model selection, allowing the Agent system to automatically split tasks, execute them in parallel, and summarize the results. This transformation accelerates ChatGPT's evolution from a chat tool to a workflow platform.
By automating task decomposition and model scheduling, these updates position ChatGPT as a more powerful and seamless interactive tool, ready to handle complex tasks with minimal user intervention.
Key facts
- ChatGPT loading speed increased by 94%.
- Memory usage reduced by 41.2%.
- Network requests dropped by 98.2%.
- Conversation history loading time reduced by 99.6%.
- GPT-5.6 Multi-Agent v2 allows automatic task delegation among different models.
- The update was tested on a 741-round conversation scenario.
- Average opening time for extreme conversations dropped from 27.62 seconds to 1.66 seconds.
- Memory growth reduced from 1030.7 MiB to 606 MiB.
Context
This performance upgrade follows the departure of Scott Gray, who was instrumental in improving ChatGPT's front-end performance. The updates address long-standing user concerns about lag and crashes during extended conversations.
The introduction of GPT-5.6 Multi-Agent v2 marks a significant step towards automating complex tasks, reducing the need for manual model selection, and lowering reasoning costs. This positions ChatGPT as a more efficient and powerful tool for developers and users alike.
