OpenAI to Pool ChatGPT and Codex Usage, Ending Decoupled Billing
OpenAI will pool ChatGPT and Codex usage for Plus, Pro and Business plans in December, ending a decoupled billing perk; how many tokens each plan includes remains unstated.
OpenAI will begin sharing usage allowances across ChatGPT and Codex for Plus, Pro and Business plans in December, measuring consumption in tokens as agentic Work capabilities move into the main ChatGPT experience. More advanced models and complex tasks generally draw more of the allowance, and OpenAI says it is preparing carefully to prevent disruption.
Key takeaways
- Instant chat will not draw from the shared usage.
- Combined remaining usage and reset times appear in Settings → Usage.
- At the limit, subscribers can wait for reset, upgrade, or buy credits for eligible features.
- OpenAI expects most users will not need to wait for resets more often than they do now.
What people think
Analysis details
9 comments analyzed. Analyzed .
The sample selects active discussion branches and includes available parent comments. It may omit parts of the full thread. The analysis included 9 of 11 usable stored comments because of the input limit. Selected themes do not measure community opinion or how common a view is.
Decoupled chat and coding budgets end
A shared allowance removes a feature that kept heavy Codex or mathematical work from consuming everyday chat and search headroom. Separate pools were a reason for loyalty; with one pool, a long run can exhaust the other workload, making reassessment or switching more likely if limits bite. Instant chat staying outside the pool is a partial mitigation, but it does not protect heavy Codex or Work users.
Token quotas remain hard to audit
Token-based limits invite scrutiny because the plan allowance is not defined. Open-source tokenizers and per-model tokenizer disclosures let users count text, but they do not answer how many tokens a Plus subscription gets per five-hour or weekly window, or whether that quota changes over time. The distinction matters: counting tokens is possible, mapping them to a subscription allowance is not.
Cheap alternatives lower switching costs
Because most project files live on disk, moving providers is not inherently costly. A $10 OpenRouter balance and DeepSeek models are proposed as inexpensive experiments, and open models serve as a hedge. The practical caveat is that a cheaper model must still fit the workflow; the cited spend is a self-reported anecdote, not a benchmark.
Parallel plan cuts test competition
A reported reduction in Claude Max from 20x to 10x, arriving around the same time as OpenAI's merge at the same price, is read as providers tightening together rather than competing on value. The comparison is secondhand and plan details may differ, but it sharpens the concern that subscription economics are moving against heavy users.
Undefined allowances raise consumer-law questions
One line of criticism argues that selling a subscription whose token allowance is undefined, unmeasurable in practice and subject to unilateral change raises consumer-protection problems, particularly in the EU. The objection is about the legality and fairness of the offer, not merely its price, and it remains a normative claim rather than a resolved legal finding.