Yes, Gemini has a limit, and it isn’t a fixed number of prompts. Since May 2026 the Gemini app uses compute-based usage limits that “factor in the complexity of your prompt, the models and features you use, and the length of your chat.” Your limit refreshes every 5 hours until you reach a weekly limit. Paid Google AI plans multiply the free allowance, and each plan also has its own context window.
Here is how the limit works, how much each plan gets, and what happens when you hit it. Plan-by-plan details are in our Gemini free limits, Google AI Pro and Google AI Ultra listings.
On this page
How Gemini limits work
Google announced the change at I/O 2026: “We’re moving from daily prompt limits to a ‘compute-used’ model.” The reasoning is that “a simple text prompt uses far less compute than a complex video or coding prompt.”
In practice that means three things:
- Every prompt costs a different amount. A short question uses little. A long chat, a big file, video generation or Deep Research uses much more.
- There are two windows. Your usage refreshes every 5 hours, but only until you reach the weekly limit.
- Settings change the cost. Google says “more advanced models and higher thinking levels consume more of your usage” (source).
Google also warns that limits “may change without notice, including due to capacity constraints.”
Gemini limits by plan
Google’s limits page expresses each plan as a multiple of the free allowance:
| Plan | Usage limit | Context window |
|---|---|---|
| No AI plan (free) | Standard limits | 32K tokens |
| Google AI Plus | 2x standard | 128K tokens |
| Google AI Pro | 4x standard | 1 million tokens |
| Google AI Ultra | 5x or 20x AI Pro, by tier | 1 million tokens |
All four tiers can use Gemini 3 Flash-Lite, Flash and Pro, according to the same page. Deep Think, the maximum reasoning level, is only on AI Ultra and requires the Pro model.
On the subscriptions page, Google AI Ultra costs $99.99 a month for 5x the AI Pro limits or $199.99 a month for 20x. For AI Pro pricing, see our Google AI Pro listing.
What uses the most of your limit
Google names the heavy hitters directly: “Some features (like media generation or Deep Research) will consume more of your usage” (source). The other factors are:
- The model you choose, since more advanced models use more
- The thinking level, since higher levels use more
- The length of the chat
- How complex the prompt is
If you’re on the free tier, Google adds that compute-heavy features such as Deep Research “may be unavailable during periods of high demand”, and free limits may be cut before paid ones when capacity changes.
The context window is a separate limit
The usage limit controls how much you can do over time. The context window controls how much Gemini can read at once. Google calls it “the reading capacity.”
Without a plan it’s 32K tokens; AI Plus gets 128K and AI Pro and Ultra get 1 million. Google says a 1M window lets Gemini understand “up to 1,500 pages of text or 30,000 lines of code.” When you go past the window, responses may miss details scattered through large files. To gauge the length of a text, try our token counter, though Gemini uses its own tokenizer.
What happens when you reach the limit
- Watch for the warning. Gemini notifies you when you’re close to your limit, then again when you reach it, with the time it will refresh.
- Check your usage. On gemini.google.com, select Settings at the bottom left, then Usage Limits (steps).
- Keep going on Flash-Lite if you subscribe. With a Google AI plan, you can continue your conversation with Flash-Lite after the limit.
- Wait or upgrade. If you reach the five-hour or weekly limit, you can wait for the refresh or move to a plan with higher limits.
Google’s subscriptions page also says you can extend Gemini app limits by buying AI credits. Back in May, Google said top-up credits were for AI Pro and Ultra subscribers.
If Gemini stops responding for everyone and you haven’t seen a limit notice, check Google’s AI Studio status page instead.
