OpenAI reinstated a 5-hour usage limit for ChatGPT Plus and Business standard users. The change appeared in a Hacker News thread that reached 120 points and 133 comments within hours.
The policy caps total active usage time rather than token counts. Users who hit the limit must wait for the rolling window to reset before continuing.
What the Limit Actually Covers
The restriction applies to standard Plus and Business plans. It measures cumulative session time across web, mobile, and API access tied to the same account. Enterprise and Team plans remain unaffected.
OpenAI previously tested similar time-based caps before shifting to token-based metering. The return to a 5-hour ceiling reverses that approach for non-enterprise tiers.
HN Community Reaction
Commenters focused on three points. Several users reported hitting the limit during normal daily workflows. Others noted the change arrived without prior email notice. A third group compared the policy to earlier OpenAI experiments that were later rolled back.
The thread shows consistent frustration with the sudden reintroduction rather than surprise at the existence of limits.
How OpenAI Limits Compare
| Service | Limit Type | Plus/Business Cap | Reset Window |
|---|---|---|---|
| OpenAI | Time-based | 5 hours | Rolling |
| Anthropic | Token-based | ~500k tokens | Daily |
| Google Gemini | Token-based | ~1M tokens | Daily |
OpenAI's time metric is stricter for heavy interactive use. Token-based systems from competitors allow longer sessions if output stays concise.
Who Feels the Change Most
Power users running extended research sessions or iterative coding assistance hit the cap first. Casual users averaging under two hours per day see little difference. Teams on Business plans now face clearer pressure to upgrade to Enterprise or spread workloads across multiple accounts.
Developers relying on the ChatGPT interface for prototyping report the largest workflow disruption.
Practical Workarounds
Users can switch to the API with separate billing to bypass the chat interface limit. Some split tasks between ChatGPT and Claude or Gemini. A smaller group moved to local models for non-sensitive work.
No official migration path or grandfathering was announced.
Bottom line: The 5-hour cap reintroduces friction for mid-tier users without improving output quality or speed.
OpenAI's decision signals tighter capacity management ahead of next model releases.
Top comments (0)