AI BriefingChatGPT, Claude, and Gemini adopt metered usage limits without publishing per-message costs
·2026.11.05 19:06
Key point
ChatGPT, Claude, and Gemini now use metered usage limits, but none publishes the specific cost of individual messages or the total allowance size.
Details
- ChatGPT, Claude, and Gemini have shifted to metered usage models. ChatGPT and Claude share allowances across specific products (ChatGPT/Codex and Claude.ai/Claude Code respectively), while Gemini uses compute-based limits.
- Claude and Gemini limits reset or refresh every five hours, while ChatGPT displays reset times without specifying a fixed interval in the provided text.
- Vendors do not publish the absolute size of the standard allowance or the specific token/compute cost of a single message. Users must rely on relative indicators like "4x higher" or percentage gauges.
- Usage consumption varies based on conversation length, complexity, and features like Deep Research. Users must manually measure usage by noting percentage changes before and after tasks, as no vendor provides a pre-run estimate for message costs.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Anthropic restricts Claude from reverse-engineering video games
Claude Recovers Ransomware-Encrypted Files by Reconstructing Data from Untouched Segments