What's New in Bots: Nov 2, 2026
OpenAI says ChatGPT and Codex will share one usage allowance starting in December, and dots are getting more models to run on, so here is what is new in bots.
ChatGPT: one allowance for chat and Codex
OpenAI updated its help article on ChatGPT and Codex usage limits over the weekend. Starting in December, Plus, Pro and Business plans will share one usage allowance across ChatGPT and Codex, as OpenAI brings the agentic side of Work into the main ChatGPT experience. Usage will be measured in tokens, so how fast it goes depends on the model you pick and the task you give it. Bigger models and harder jobs use more. Instant chat will not count against it at all.
OpenAI has not given an exact start date or new allowance sizes. It says it expects most people will not wait on a reset more often than they do now, and you will see your combined remaining usage and reset times under Settings, then Usage. If you run out, you can wait for the reset, upgrade your plan, or buy credits.
What it means for you: a big Codex session or a long running task will come out of the same pot as your everyday chats. Running a new job as a batch of ten first is a cheap way to see what it really costs before you scale it up. (The doubled Work and Codex usage on Pro was in our Oct 30 roundup, and account switching on mobile was in Saturday's issue.)
Dots: more models on the way
Tibo, who works on Codex and ChatGPT at OpenAI, posted this morning that OpenAI is working on expanding the family of frontier models that power dots, including GPT-6 Astra and GPT-6.1 Terra, with others to come. He says Terra brings speed and "a nice personality upgrade," and will help OpenAI bring dots to more people. Dots launched on GPT-6 Astra, per the ChatGPT release notes. There are no dates or plan details yet, so treat this as a heads up, not a launch. (Texting your dot was in our Oct 29 roundup.)
DeepSeek: a prototype team coding agent
DeepSeek open sourced an experimental agent called Super-Agent, under the MIT license. You point it at one GitHub repository (and, if you like, Feishu, ByteDance's work chat). When someone you allow asks it for work, it opens a work item, does the job in its own session, and answers in the original thread, for example with a pull request. A web panel shows what needs a person. DeepSeek calls it a "research prototype for reference only, not an official DeepSeek product," and says to review and adapt it before you use it. As always, check the license and try it on a side repo first, per our guide to using open source with your bots.
Quick Facts
- From December, Plus, Pro and Business share one token based allowance across ChatGPT and Codex.
- Instant chat does not count against it, and there is no exact start date or allowance size yet.
- OpenAI is working on running dots on more models, including GPT-6.1 Terra.
- DeepSeek open sourced Super-Agent, a prototype agent that takes work from GitHub threads.
- Want to see how these bots stack up? Run your own bot vs. bot bake off or browse the directory.
See you tomorrow.