Gemini's free tier drops to Flash-Lite only on October 9

Starting October 9, 2026, people who use the Gemini app without a paid Google AI plan can only use Flash-Lite, Google's smallest model. Flash and Pro, the faster and stronger models, move behind paid plans. The change is set out on Google's help page "Changes to Gemini model access and limits." It narrows Google's free AI assistant a day after ChatGPT moved its free users to GPT-6.

The cheapest paid plan changes too. AI Plus, which costs $4.99 a month in the U.S., loses access to Pro, iThinkDifferent reports. AI Plus subscribers get an email with the date the change reaches their account.

Which plan gets which model

Google sells three paid tiers above the free app. After the change, access looks like this, according to Google's help page and Android Headlines:

Plan	Flash-Lite	Flash	Pro	Deep Think
No plan (free)	Yes	No	No	No
AI Plus ($4.99 a month)	Yes	Yes	No	No
AI Pro ($19.99 a month)	Yes	Yes	Yes	Yes
AI Ultra	Yes	Yes	Yes	Yes

Flash-Lite is the smallest of the three models. Deep Think is an extra reasoning option, which Google's page describes as "maximum parallel reasoning," as quoted by DigitBin. An older Google help article listed Deep Think for AI Ultra only, DigitBin reports, so AI Pro subscribers gain it.

Prices come from iThinkDifferent. Google's help page does not list them.

Who is affected and when

The new rules cover personal Google accounts, iThinkDifferent and DigitBin report. Google's page does not cover work or school accounts. None of the sources say anything about the Gemini API, the paid service developers call from their own code.

The dates differ by plan:

• Free users: the change starts on October 9.
• AI Plus subscribers: each person gets an email with their own date. Google has not published one date for everyone.
• AI Pro and AI Ultra subscribers: they lose no models.

Google also recommends updating the Gemini app to get the new options, iThinkDifferent reports.

How the usage limits work

Gemini's limits are based on how much computing each request uses, not on a simple message count. Google introduced these limits on May 17, 2026 for users over 18 and on July 24, 2026 for users under 18, its help page says. The allowance refreshes every 5 hours until a weekly cap is reached.

Paid plans get bigger allowances. Compared with the standard limit, AI Plus gets 2 times and AI Pro gets 4 times as much. AI Ultra gets 5 or 20 times the AI Pro limit, depending on the tier, according to Google's page.

Some features use up the allowance faster. Google lists media generation, Deep Research, the Pro model, extended thinking and Deep Think. Each model also gets low, medium and high effort settings, and higher effort uses more of the limit.

Reports disagree on the exact model versions behind the names. Google's page calls them only Flash-Lite, Flash and Pro.

What this means for developers

• Expect weaker answers on free accounts. If you use the free Gemini app for coding help or quick research, you now get Flash-Lite. Test it on your usual questions before you rely on it.
• Pick the cheapest plan that keeps your model. Flash needs AI Plus at $4.99 a month. Pro and Deep Think need AI Pro at $19.99.
• Check your API plan separately. This change covers the Gemini app on personal accounts. Nothing in the sources says API access changes, so confirm with your API pricing page.
• Use effort settings to stretch your limit. Low effort saves allowance for routine tasks. Keep high effort and Deep Think for hard problems.
• Compare free tiers again. Google is narrowing its free tier just as OpenAI widens ChatGPT's. Teams that standardized on one free assistant may want to re-test both.

AI Plus users should watch their inbox for the date Pro disappears from their account.