Google is no longer keeping its newest Gemini model behind its priciest subscription wall. Gemini 4 Argon, which launched in late September exclusively for $99.99-a-month AI Ultra subscribers, started rolling out Monday to the cheaper Google AI Pro tier inside the Gemini app, according to 9to5Google.
Google is pitching the update as a jump in "frontier reasoning and improved performance across complex workflows," language that mostly points at harder, multi-step tasks rather than everyday chat. AI Pro and AI Ultra subscribers will now see four models in the Gemini app's picker: 3.5 Flash-Lite, 3.8 Flash, 3.1 Pro, and the new 4 Argon. Both tiers also get Argon access inside AI Mode, Google's AI-driven search layer.
The bigger changes land on macOS. Because Argon brings what Google calls "major improvements" to coding, the Mac version of the Gemini app now lets users flip between chatting and coding without leaving the client. That switch hands users direct access to the agentic coding interface from Google Antigravity, the company's developer tool, right inside the consumer app, so building a simple web page or app no longer means jumping to a separate program.
There's also a new toggle in the main chat window for switching between chat and tasks. Tasks is the renamed version of what Google previously called Spark.
Ultra still gets more
Argon's spread to AI Pro has a ceiling. The Antigravity desktop app and its command-line tool still require an AI Ultra subscription to use the new model, as do AI Studio and the Canary build of Android Studio. Google used the rollout to show off internal uses of Argon inside its own version of Antigravity, pointing to work across Chrome, its Gemma model family, and YouTube, though none of those are user-facing features yet.
For developers building on top of Argon rather than using Google's own apps, the Gemini API is charging an introductory rate of $2 per million input tokens and $10 per million output tokens. That pricing holds until November 30, 2026, after which it doubles to $4 per million input tokens and $20 per million output tokens. Tokens are the chunks of text a model reads and generates, and per-token pricing is how most AI labs bill developers for API use.
Google also opened a separate program aimed at researchers rather than developers. It's offering 10,000 scientists free access to the "highest usage tier" in Google Antigravity for six months, with applications now open, though the company hasn't said how recipients will be chosen or how long the application window stays open.
The net effect is a faster-than-usual walk-back of an Ultra-only launch window, something Google has done before with flagship Gemini releases as it tries to keep AI Pro, its $19.99-a-month tier, competitive against ChatGPT Plus and other rivals. Whether Argon reaches the free tier of Gemini, and when, Google hasn't said.
ADFiled by The AI Desk
Models, assistants and the companies and chips behind them, reported from what was released and what was claimed, with the difference kept clear.
More from this desk →
Be the first to comment
Join the argument. No password, just your email or a passkey.