OpenAI 12月共享的ChatGPT和编码使用津贴

中文版正在随流水线自动整理，稍后刷新即可；可先阅读下方英文原文。

本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。

AI Models & Platforms

OpenAI Sets Shared ChatGPT and Codex Usage Allowance for December

Add Unite.AI to your preferred sources on Google

OpenAI will move Plus, Pro, and Business subscribers to a single usage allowance shared across ChatGPT and Codex starting in December 2026, according to a notice published on the company’s Help Center. Instant chat will not draw from the shared pool.

Under the new structure, usage for ChatGPT and Codex will be measured in tokens, the units of information a model processes and generates. OpenAI said how much of the allowance a subscriber consumes will depend on the model chosen and the task requested, with more advanced models and more complex tasks generally using more of it. The company framed the change as part of bringing the agentic capabilities of Work — its ChatGPT experience built for longer research and deliverable tasks — into the main ChatGPT experience.

Once the change is in effect, subscribers will be able to see combined remaining usage and reset times under Settings → Usage. Users who reach the limit will be able to wait for it to reset, upgrade their plan, or buy credits to continue using eligible features, according to the notice.

What the Change Consolidates

OpenAI already meters much of this activity from shared pools. On plans that use shared allowances or credits, Codex, ChatGPT Work, ChatGPT for Excel, and Workspace Agents draw from a shared allowance and credit pool, per OpenAI’s Codex documentation. ChatGPT for Word also draws from the shared usage allowance for Codex and other premium features, according to the company’s release notes.

Codex itself is included across ChatGPT plans, including Free and Go. Codex Cloud, which runs coding tasks on OpenAI-managed computers, is available to eligible Plus, all Pro tiers, Business, Enterprise, Healthcare, and Education accounts. The documentation states that consumption depends on the model, where a task runs, task complexity, context, reasoning, speed, and tools, and that a long-running task can use substantially more allowance than a short request. Codex limits run on five-hour and weekly windows, and using a full banked reset refreshes both windows while changing the weekly reset date. OpenAI Support does not reset ChatGPT or Codex usage limits.

Business Seats and Separate Chat Limits

ChatGPT Business bundles ChatGPT, Work, and Codex and is sold in Standard and Premium seats. OpenAI’s plan documentation states that Premium seats include more Work and Codex usage and carry no five-hour usage limit, while Standard seats operate with per-model local-message estimates per five-hour window: for example, 15–150 messages on GPT-6.1 Sol or GPT-6 Sol and 250–2,000 on GPT-5.6 Luna, with actual usage varying by model and task. Pro-model chat allowances remain separate from Work and Codex usage: Business Standard includes 15 Pro messages per month and Business Premium 50 per week, shared across GPT-6 Pro and GPT-5.6 Sol Pro. The documentation describes Instant chat as virtually unlimited, subject to anti-misuse safeguards under OpenAI’s Services Agreement.

Recent Adjustments to the Same Pools

The December plan follows a series of changes to the same allowances. ChatGPT release notes show that on October 29, 2026, OpenAI doubled the Work and Codex allowance included in Pro 100 and Pro 500 plans and canceled a Pro 200 reduction announced in a September 29, 2026 email, giving affected Pro 200 subscribers twice their previous allowance and bringing them to the same limit. The September 29 notes also introduced Pro 500, a $500-per-month plan with the highest included usage of the Pro plans, and reopened Pro 200 subscriptions at $200 per month.

OpenAI said it is preparing carefully for the December transition to help prevent disruption to user workflows, and that it expects most users will not need to wait for their usage to reset more often than they do now.

Auto-fetched and translated by AI Atlas. All rights belong to the original author.