ChatGPT to merge Chat and Work under one token budget

OpenAI is preparing to merge Chat and Work in ChatGPT, and the change will also reshape how usage limits work. Starting in December, Plus, Pro and Business plans will use a single token budget across ChatGPT and Codex, according to an updated OpenAI support page cited by The Decoder.

For creators, freelancers and small teams, the important shift is not just interface cleanup. It means heavier work sessions, coding tasks and regular chat use will draw from the same pool, making ChatGPT feel less like an unlimited prompt box and more like a managed resource.

One pool for chat and work

OpenAI had already said it planned to bring Chat and Work together. The new support-page detail, according to The Decoder, is that the merger will also unify usage accounting across ChatGPT and Codex for Plus, Pro and Business customers.

ChatGPT is already used by many small teams as a writing assistant, brainstorming partner, research aide and coding helper. Under the upcoming setup, those different modes of work will no longer sit behind separate-feeling usage boundaries. Instead, they will share one token allowance.

Tokens are the unit used to measure what a model reads and produces. Longer inputs, longer outputs and more demanding tasks can consume more of that budget. The source says more complex models and jobs will use up the allowance faster.

OpenAI will show users their remaining balance and reset times in Settings, according to the support page cited in the report. When a user reaches the limit, the available options are to wait for the budget to reset, move to a higher plan or purchase extra credits.

What still stays flat-rate

The change does not mean every ChatGPT interaction will count against the new pool. The Decoder reports that chats using Instant models will not count toward the token budget.

That distinction matters for day-to-day workflows. A freelancer using ChatGPT for quick rewrites, outlines or short questions may be able to keep some low-stakes work away from the budget by using Instant models. But more demanding sessions could still draw down the shared allowance.

The Decoder’s read is that ChatGPT can still be presented as a flat-rate product on paper, because Instant model chats are excluded. But the outlet also argues that heavy users on lower-priced plans may end up relying more on lower-quality results if they want to avoid burning through their allowance.

For users who treat ChatGPT as an always-open assistant, the practical habit change is simple: watch the balance. Teams should also think about which tasks really need stronger models and which can be handled through lighter chat.

Why OpenAI is simplifying ChatGPT

The merger also appears to address ChatGPT’s growing complexity. The Decoder notes that users currently have to choose between two operating modes and 11 model settings, a setup that can be confusing even for experienced users.

Folding Work’s agentic features into the main interface could make ChatGPT easier to understand, especially for teams that do not want to train staff on multiple modes. Emonarc readers who use ChatGPT for marketing calendars, client drafts, spreadsheet help or coding support may welcome fewer product boundaries, even if the new quota system requires more attention.

OpenAI has also recently doubled usage limits for higher-priced plans, according to The Decoder. That move may have been preparation for the merged system. Emonarc previously covered the related change in OpenAI doubles Pro usage limits after user pushback.

The broader pattern is that AI assistants are moving from simple chat windows toward work environments. OpenAI has been bringing more agent-style ideas into ChatGPT, including coding-related concepts covered in OpenAI’s dots brings coding-agent ideas into ChatGPT. The upcoming Chat and Work merger fits that direction.

The Claude comparison

OpenAI is not alone in collapsing chat and work features into one experience. The Decoder points out that Anthropic already combined Chat and “Cowork” in Claude.

That comparison is useful for buyers because the market is moving away from isolated chatbot sessions and toward assistants that handle more sustained tasks. As that happens, pricing and usage limits become more important than the number of prompts a user can type.

For small businesses, the key question is operational: who on the team is consuming the shared budget, and on what kind of work? If one person uses the assistant heavily for complex tasks, others may feel the constraint sooner. OpenAI’s Settings view for balance and reset times should help, but teams may still need internal norms around when to use more capable models.

The December change will test whether OpenAI can make ChatGPT simpler without making usage feel more restrictive. For now, the clearest takeaway is that heavy users should expect token budgeting to become part of their ChatGPT workflow.