OpenAI doubles Pro usage limits after user pushback

OpenAI has doubled usage limits for several of its paid Pro plans, according to The Decoder, with bigger allowances for Work and Codex quotas on Pro 100 and Pro 500. The company also changed course on Pro 200 after user pushback, while leaving Plus unchanged.

For creators, freelancers and small teams paying for higher-tier AI access, the update is less about a new feature and more about reliability. OpenAI is effectively giving some paying customers more room to run coding, work and agent-style tasks before hitting plan limits, at a time when rival AI providers are competing hard for business users.

What changed across the Pro tiers

The quota increase applies to several paid tiers rather than the whole product line. According to The Decoder, Pro 100 and Pro 500 now receive twice their previous Work and Codex allotments. On Pro 200, OpenAI doubled the lower quota and backed away from a planned reduction after customers objected.

New subscribers already get the revised quotas. Existing customers receive the increases automatically, so there is no separate opt-in step mentioned in the report.

For teams using ChatGPT as a daily workspace for drafting, coding help, analysis or internal workflows, the practical benefit is more headroom. A quota increase can reduce the need to ration prompts during busy periods, especially for users who already found themselves running into caps on Pro plans.

The Plus plan is not changing. That matters for solo users and budget-conscious freelancers: the update appears targeted at heavier paid usage, not the mainstream individual subscription tier.

Why twice the quota may not mean twice the work

OpenAI is also cautioning users not to read the change too literally. The company says that doubling a quota does not guarantee twice as many completed tasks, because usage depends on the model, the type of request and selected settings.

That distinction is important for anyone trying to estimate subscription value. A long coding task, a multi-step work request or an agent-like workflow may consume capacity differently from a short writing prompt. In practice, the same plan can feel more or less generous depending on how it is used.

A simple way to think about the change:

• Heavy users on Pro 100 and Pro 500 get twice the stated Work and Codex quota.
• Pro 200 users get a higher lower quota, and a planned cut has been reversed.
• Plus users should not expect a change from this update.
• Existing subscribers do not need to take action to receive the new limits.
• Actual output will still vary by model, task and settings.

For agencies and small businesses, this means quota planning still requires observation. If your team uses AI mostly for short marketing drafts, the extra allowance may stretch further than it would for long coding sessions or complex task chains.

Dots stays expanded, but tracked separately

The Decoder also reports that the expanded limits for OpenAI’s new Dots remain in place for now. Dot usage is tracked separately, so those limits should not be treated as the same pool as the Work and Codex quotas discussed in this update.

That separation matters for users experimenting with newer agent-style workflows. If Dots becomes part of how a team delegates tasks inside OpenAI’s ecosystem, separate tracking could make it easier to understand where usage is going. It also means teams should watch more than one meter when evaluating whether a plan is enough.

Emonarc previously covered how OpenAI’s Dots brings coding-agent ideas into ChatGPT. This quota update suggests OpenAI is still adjusting the capacity side of that experience as customers test where it fits in real work.

The subscription fight is moving to capacity

OpenAI says it will keep improving models and efficiency in order to increase usable capacity. The company recently released its 6.1 models, Terra and Sol, which it says are much more efficient while matching or outperforming earlier versions. We covered that model update separately in OpenAI’s GPT-6.1 Terra cuts API prices while improving benchmarks.

The broader timing is notable. The Decoder frames the move against rising competition with Anthropic, which is preparing for an IPO, as both companies compete for business AI customers. Anthropic’s financial and infrastructure picture has also become more visible through its IPO process, as covered in Anthropic files IPO prospectus, revealing Claude maker’s losses and compute deals.

For buyers, the signal is that paid AI subscriptions are no longer competing only on model quality. Capacity, quota transparency and how limits apply to coding or work tasks are becoming part of the purchase decision.

Small teams should treat this as a prompt to audit actual usage. If you are on Plus, nothing changes. If you are already on a Pro tier, the update may reduce friction. If you are comparing vendors, the key question is not just which model performs best on a demo task, but how many real tasks your plan can sustain during a normal workweek.