AI % min read

Google Changes Gemini Limits From Fixed Requests to Compute Based Usage

Google Changes Gemini Limits From Fixed Requests to Compute Based Usage
Photo by Mitchell Luo / Unsplash

Google is shifting Gemini from fixed daily request limits to a compute‑based system that measures usage by prompt complexity, features used, and chat length. Paid tiers now scale significantly, with AI Plus offering 2x standard limits, AI Pro 4x, and AI Ultra 20x, while limits refresh every five hours under a weekly cap. The change mirrors GitHub Copilot’s recent move to token‑based billing as advanced agentic AI features consume far more compute. As AI models spawn sub‑agents and handle longer reasoning chains, flat‑rate plans are becoming unsustainable for providers.

Read the full story on PCWorld →