What is happening to free Gemini?
Starting October 9, 2026, Google is restricting which Gemini models free users can access in the Gemini app. According to an updated Google support page first reported by 9to5Google on October 3, anyone using Gemini without a Google AI subscription will be limited to the 3.5 Flash-Lite model. Access to the stronger 3.6 Flash and 3.1 Pro models will be removed for free accounts.
The paid tiers are shifting too. Google AI Plus subscribers paying $4.99 a month will keep Flash-Lite and Flash but lose access to Pro. Google says Plus subscribers will receive an email explaining exactly when the change takes effect for their account, so the cutover date may vary from person to person.
The one clear upgrade lands on the middle tier. AI Pro at $19.99 a month gains Deep Think, a maximum parallel reasoning mode that was previously reserved for the $99.99 and $199.99 AI Ultra plans. AI Pro and AI Ultra keep access to all three models with no removals.
Which Gemini model does each plan get after October 9?
Here is the breakdown once the change rolls out:
- No subscription (free): Flash-Lite only. Flash and Pro are removed on October 9.
- Google AI Plus ($4.99/month): Flash-Lite and Flash. Pro is removed, with timing communicated individually by email.
- Google AI Pro ($19.99/month): Flash-Lite, Flash, and Pro, plus the new Deep Think option.
- Google AI Ultra (from $99.99/month): Flash-Lite, Flash, Pro, and Deep Think. No changes.
Flash-Lite is the fastest and lightest of the three models. Flash sits in the middle for everyday work. Pro is the most capable model for complex writing, coding, and reasoning tasks. For free users, the practical effect is simple: after October 9, Gemini becomes noticeably less capable on longer or harder prompts.
Why Google is shrinking the free tier
The change follows Google’s switch in May 2026 from fixed message counts to usage limits based on computing cost. A heavy model like Pro burns through that allowance far faster than a light one. Removing heavy models from the free tier is a blunt but effective way to control inference costs.
It is also a sales funnel. Flash-Lite handles quick questions just fine, which keeps free users in the ecosystem, while anyone who regularly needs stronger answers will feel the ceiling and look at the paid plans. Google framed the change on its help page as a routine update to model availability, not a downgrade, but in practical terms free users lose two of the three models they can choose today.
This is the second time in 2026 that Google has redrawn the line between what casual users get and what subscribers pay for, and it lands in the same week Google added Guided Vision to Gemini Live.
New effort settings are coming too
Alongside the tier reshuffle, the Gemini app is expected to add low, medium, and high effort levels for each available model in October. According to Google’s documentation, these settings will increase the model’s ability to complete tasks and provide more thorough answers to queries, but they will also use more of your usage limit.
Today, the closest equivalent is the existing “Extended thinking: Complex problem solving” toggle. The three new effort tiers would bring the consumer app in line with controls already offered in Google AI Studio and Antigravity. One open question is where Google’s next-generation model, Gemini 4 Argon, fits in. Google has said Argon is first coming to AI Ultra subscribers, but it has not specified whether it will be classified as a Pro model or sit in a tier of its own.
Does this change how Gemini compares to ChatGPT?
For people comparing assistants, the October 9 change makes the comparison more interesting. ChatGPT’s free tier still offers limited access to more capable models, with caps that vary by demand and feature rather than a hard lockout to a single model. Gemini’s new structure is stricter: free users get exactly one model, full stop.
If you already compared the two assistants for work, this is a good moment to reassess. Our Claude vs ChatGPT breakdown for small businesses walks through pricing and capabilities across the major assistants, and Meta’s own AI assistant offering shows how crowded the field is getting.
What should free users do?
You have three practical options:
- Stay free and adapt. Flash-Lite is fast and fine for quick questions, summaries, and everyday lookups. If that covers most of your usage, nothing forces you to pay.
- Watch for the Plus email. If you pay $4.99 a month for AI Plus, note that your Pro cutoff may not land exactly on October 9. Google will email you your specific date.
- Consider whether AI Pro earns its price. At $19.99 a month, AI Pro now keeps all three models and adds Deep Think, a feature that previously required the $99.99 Ultra plan. For people who use Gemini for research, writing, or coding, that is a materially better offer than it was last month.
The bottom line: Google is narrowing what free AI means inside the Gemini app. If you want the strong models, October 9 is the date that matters, and AI Plus is no longer a complete escape hatch.
Frequently asked questions
When do the Gemini free tier changes take effect?
October 9, 2026, for users without a Google AI subscription. Free users lose access to Flash and Pro on that date and keep only Flash-Lite. AI Plus subscribers get an individual effective date by email later in October.
Which Gemini model will free users still be able to use?
Gemini 3.5 Flash-Lite, Google’s smallest and fastest model. Selectable access to 3.6 Flash and 3.1 Pro goes away for free accounts.
Does Google AI Plus still include Gemini Pro after October 9?
No. AI Plus at $4.99 a month keeps Flash-Lite and Flash but loses Pro access, with the exact timing communicated to each subscriber individually.
Which Google AI plan keeps full access to every Gemini model?
Google AI Pro at $19.99 a month and Google AI Ultra at $99.99 a month both retain Flash-Lite, Flash, and Pro. AI Pro also gains the Deep Think feature in this update.
Is this related to the usage limits Google introduced earlier in 2026?
No. In May 2026, Google switched Gemini’s usage accounting from simple prompt counts to compute-based limits. The October 9 change is separate: it restricts which models each tier can even select, regardless of how much usage you have left.
Does this affect Gemini on business, school, or enterprise accounts?
The restrictions reported so far apply to personal Gemini accounts. Organizational and enterprise plans are tiered separately, so check with your account administrator for specifics.
What are the new low, medium, and high effort settings?
Google is expected to add per-model effort levels to the Gemini app in October. Higher effort should produce more thorough answers for harder tasks but consume more of your usage limit. They have not launched yet.