ChatGPT to share one token budget across Chat and Codex on Plus, Pro and Business from December

Starting in December, ChatGPT Plus, Pro and Business plans will share a single token budget across ChatGPT and Codex, according to an updated OpenAI support page. The change accompanies OpenAI's plan to fold Work's agentic features into the main ChatGPT interface, as first reported by The Decoder on Nov 1, 2026.

OpenAI had already announced plans to merge Chat and Work.

How the new limits work

Tokens measure the information a model processes and generates. More complex models and tasks consume the budget faster.

• Shared budget: Plus, Pro and Business plans draw from one token pool covering both ChatGPT and Codex.
• Visibility: Users can check their remaining balance and reset times in Settings.
• At the limit: Users can wait for a reset, upgrade their plan, or buy extra credits.
• Instant models: Chats with Instant models do not count toward the budget.

The size of the token budget for each plan is not yet disclosed.

Context: limits and interface complexity

OpenAI recently doubled usage limits for its higher-priced plans. The Decoder presumes this was preparation for the merger. The Decoder also describes the merger as likely a response to criticism of ChatGPT's growing complexity. Users currently choose between two operating modes and 11 model settings, differences The Decoder says even experienced users can barely keep straight.

Anthropic has already combined Chat and "Cowork" in Claude, according to The Decoder.

The Decoder's assessment

Because Instant model chats are excluded from the budget, ChatGPT stays a flat-rate service on paper, The Decoder notes. It adds that heavy usage on cheaper plans will likely mean lower-quality results, since users who exhaust their budget on more capable models would have to fall back on Instant models or wait for a reset. That is The Decoder's analysis, not an OpenAI statement.

What this means

The change moves ChatGPT's paid plans from separate usage allowances toward a single metered pool. A long agentic task in Codex and a complex chat will now draw from the same balance, so the choice of model and task type directly affects how long a subscription lasts. Instant model chats remain outside the meter, which keeps a free-to-use tier of interaction inside paid plans. For users on Plus and Business, the practical questions are how fast their typical workload burns tokens and whether the Settings balance display gives enough warning before a limit hits. OpenAI has not yet disclosed the per-plan budget sizes, which are the figures needed to judge that.