Launching today
Gemini 4 Argon
Google's frontier model for long, complex agentic work
33 followers
Google's frontier model for long, complex agentic work
33 followers
Gemini 4 Argon is Google's frontier model for long horizon work in software engineering, enterprise knowledge work and cybersecurity defense. As of Oct 13 it is available globally in the Gemini API paid tier, AI Studio, Google Antigravity (Ultra), the Gemini app and AI Mode in Search (Pro and Ultra), with a 1M unified input and output token limit.
Launch Team
LightfieldAI-native CRM that builds itself and does work for you
Promoted
Hey PH community 👋
Excited to hunt Gemini 4 Argon today.
Google just made Argon, their new frontier model, available globally, and they say it is built for long horizon work in software engineering, enterprise knowledge work and cybersecurity defense. It's live for developers in the Gemini API and for Pro and Ultra subscribers in the Gemini app and AI Mode in Search. Sundar posted the launch on X here.
Here's some examples from the community:
https://x.com/thtbee_/status/2110044715618116000
https://x.com/tinypia/status/2110044780629799290
The 1M token limit is the part I'm most excited about. Google says Argon gets a unified 1M input and output limit in the Gemini API, which means it can keep working through a big code migration or research task instead of stopping halfway and asking you to start over.
Long tasks in the Gemini app are the other thing I'd try first. Google says Argon answers more directly with less hedging, and it keeps you posted on its progress while it works through something bigger.
For builders, it's in the Gemini API paid tier today, and Ultra subscribers can also use it in Google Antigravity and AI Studio.
You can try it in Google AI Studio right now: https://aistudio.google.com
Give it a spin and let the team know what you create with Gemini 4 Argon 🚀
PS: Tagging @logankilpatrick and @ammaar if you're here for any additional comments/ thoughts.