Gemini app deploys Gemini 4 Argon on AI Pro and AI Ultra

Following the announcement in late September, Gemini 4 Argon is starting to roll out to the Gemini app, Google Antigravity, and more.

Gemini app

In the Gemini app, Gemini 4 Argon is available to Google AI Pro and AI Ultra subscribers. Last month, initial availability was only for the Ultra tier ($99.99 and up).

Google says to expect “border reasoning and improved performance in complex workflows” in the app. In addition:

• “Argon answers questions more directly with less coverage, doesn’t repeat himself more than necessary, and avoids clichés throughout the answers.” Plus, it “follows styling instructions more reliably than any of our previous Gemini models.”
• “For longer tasks, it keeps you in the loop by sharing its progress more often and more effectively.”

Given the “major improvements” in coding performance, Google is allowing users to “switch between chat and coding” within the Gemini app for macOS. Users will specifically have direct access to Antigravity’s agent coding interface in the consumer client to create applications, web experiences, and more.

Advertising – scroll for more content

Meanwhile, there will be a new toggle in the main chat window that will allow you to switch from chat to Tasks (formerly Spark). Starting today, AI Pro and AI Ultra subscribers will see 3.5 Flash-Lite, 3.8 Flash, 3.1 Pro, and 4 Argon in the model selector.

AI Pro and AI Ultra subscribers can also access Gemini 4 Argon in AI mode.

Google Antigravity and more

In the Antigravity desktop app and Antigravity CLI, the new model requires AI Ultra. Google today shared more examples of how it’s using Argon with its internal version of Antigravity on Chrome, Gemma, YouTube and more:

• “Argon helped the Chromium The team separates management of core web pages from a legacy folder of approximately 6 million lines of browser functionality contained in Chromium’s 40 million line code base. By making several clean code changes at once, Argon helped the team complete an architectural refactoring on hundreds of files, turning months of work into weeks. For the billions of Chrome users around the world, this will allow live web pages to move seamlessly between tabs, side panels, and floating windows without reloading, eliminating an entire category of browser crashes.
• “Using Argon, we achieved a 2.85x end-to-end inference speedup on an A100 via vLLM (accelerating throughput from approximately 37,900 to over 108,000 tokens per second on an A100). Gem 4 E2B) with 100% token parity and zero quality drop in just a few days, a process that would typically take months. View upstream pull requests (optimizing everything from attention tuning to speculative decoding) to see the exact changes.
• “Argon in Antigravity helped the team reduce the CPU footprint of YouTube frontend comments in two. By plotting production profile data, Argon identified that redundant emoji rendering alone accounted for nearly 45% of the server overhead, leading to 12 targeted optimizations. The updates reduced server comment loading latency by 30ms and client end-to-end latency by 20ms.

AI Ultra is also required for access to Argon in AI Studio and Android Studio (Canary).

Argon is available in the Gemini API at an introductory price of $2 per million input tokens and $10 per million output tokens until November 30, 2026. Thereafter, it will cost $4 per million input tokens and $20 per million output tokens.

On the scientific side, Google is offering 10,000 scientists “the highest level of use of Google Antigravity free for 6 months”. Applications are now open.