Google limits free Gemini users to Flash Lite from Oct. 9 and removes Pro from the $4.99 AI Plus plan

Google will limit free Gemini users to the Flash Lite model starting October 9, according to The Verge. Free users can currently choose between Gemini Flash Lite, Flash, and Pro. After the change, standard Flash will require a paid plan.

What is changing

• Free plan: Restricted to Flash Lite only from October 9. Flash and Pro access ends.
• Google AI Plus ($4.99/month): Becomes the entry point for standard Gemini Flash. The plan will no longer include Gemini Pro.
• Google AI Pro ($19.99/month) and AI Ultra ($99.99/month): The only plans with access to the top-tier Gemini Pro model and the "Deep Think" advanced reasoning option.

Google says AI Plus subscribers will receive an email telling them when they will lose access to Gemini Pro. The exact date for that change has not been disclosed.

Effort levels and usage limits

Google says users will still be able to choose low, medium, or high effort levels for each model. According to Google, higher effort levels may consume a user's Gemini usage limit faster. The specific usage caps per plan have not been disclosed.

What has not been disclosed

• The date AI Plus subscribers lose Gemini Pro
• Numeric usage limits for any tier
• Whether API or Google AI Studio free-tier access is affected. The report covers only the consumer Gemini app plans.
• Context window sizes and per-token pricing for Flash Lite, Flash, and Pro in this announcement. None were provided.

No new model is being released. This is a change to plan access for existing models.

What this means

This is a tier consolidation rather than a model update. Google is moving each model one rung up its subscription ladder: Flash goes from free to $4.99, and Pro goes from $4.99 to $19.99. For AI Plus subscribers, the plan's value drops without a price change. The cheapest route to Pro now costs four times as much.

The linkage between effort levels and usage limits matters. Letting users select high effort on every model while drawing down a shared allowance means heavy reasoning use will exhaust quotas faster, so the effective capacity of each plan will depend on how users configure it. Without published limits, subscribers cannot easily compare plan value.

The move also signals pressure on inference costs for free consumer access. Restricting the free tier to the lightest model reduces compute spend on non-paying users and pushes engaged users toward paid conversion. Developers and teams that rely on the free Gemini app for Flash-class or Pro-class output should check whether their workflows need a paid plan before October 9.