Google's Gemini tier cuts land on 9 October, as speech and translation models keep shipping
Google will cut free Gemini users to Flash-Lite from 9 October, according to an updated support document reported by 9to5Google, while its more capable Flash and Pro models move behind paid plans.

Google will cut free Gemini users to Flash-Lite from 9 October, according to an updated support document reported by 9to5Google. The change removes access to 3.6 Flash and 3.1 Pro for anyone without a Google AI subscription.
Why it matters to anyone who uses AI for speech or translation: the models behind a free voice assistant are about to change, and the better tiers cost money. The AI Plus plan at $4.99 a month keeps Flash-Lite and Flash but loses Pro, 9to5Google reported on 3 October. Only AI Pro at $19.99 and AI Ultra at $99.99 or $199.99 keep all three. Google has not said how many users sit on each plan.
The Decoder, which reviewed the same support page, put the practical effect in blunt terms: most casual users have no idea which model they are running, and the power users who care are probably already paying. The site also noted that Google may be clearing room for Gemini 4 Argon, which could cost more to run. Where Argon fits is still open. 9to5Google reported that Google said Argon would come first to AI Ultra subscribers, without saying whether it counts as a Pro model or a tier above.
AI Pro subscribers do get something back. 9to5Google reported that the $19.99 plan will gain the Deep Think option for "maximum parallel reasoning," which until now has sat on the $99.99 and $199.99 plans. The Gemini app also looks set to add low, medium and high effort levels for each model.
Speech and translation work is where the tier line bites hardest, because those tasks lean on the larger models. Voice features have been a growth area across the sector: Anthropic now asks Claude users to let it train on their voice conversations, BleepingComputer reported on 4 October. The prompt is optional and off by default, and users can delete the data later in settings, the report said. That opt-in is separate from the toggle covering chats and Claude Code sessions.
The model plumbing underneath all this is getting more specialised too. Ollama's 0.35.1 release added Cloudflare's Clef and Clef Flash, classifiers that answer typed questions instead of writing text, according to a write-up by Jerry Gamblin, who serves on CVE Program working groups. Running Clef Flash locally, he found that 989 of 23,736 CVE descriptions outside the Linux kernel, 4.2%, never state a security impact. The 9B model answered two questions per record in under a second each on a laptop, with no API key.
Training data is the other front. OpenAI said on 4 October that it had disrupted a coordinated campaign to extract protected reasoning from its models, with activity first seen in the first week of July. OpenAI attributed a core cluster to individuals associated with Moonshot AI, the developer of Kimi, and said it shared details through the Frontier Model Forum. It is unclear whether all the operators were one actor, the company said.
For Gemini users, the date to watch is 9 October. 9to5Google reported that AI Plus subscribers will get an email explaining when the changes take effect. Google has not published a date for Argon's wider availability.
Sources
8- 01Gemini app limiting what models free and AI Plus users can accessEN
- 02Google's new Gemini tiers cut free users to its weakest model and lock $5/month subscribers out of ProEN
- 03Anthropic asks Claude users to share voice data for AI model trainingEN
- 04Decision Models vs. CVE DataEN
- 05Disrupting a coordinated model-distillation campaignEN
- 06OpenAI's latest features take direct aim at the app store modelEN
- 07Language Model "Shape"EN
- 08No Model Required: Text Entropy Rate Filtering Mitigates Iterative Fine-Tuning CollapseEN
All figures and quotations in this text come from the sources listed below.
Content prepared by the editorial team with AI assistance.
Comments
0- No comments yet — be the first.