Key points
- Argon is open to API users now; the Gemini app and Search AI Mode get it for Pro and Ultra plans, Antigravity and AI Studio for Ultra only, per Ars Technica.
- Intro API rates are $2 for input and $10 for output per million tokens, and per Ars Technica they double after November 30.
- Google says Argon ranks first on the Vals Index and cites internal gains, including 35% more agentic code submissions in Antigravity.
- Google is offering 10,000 scientists free top-tier Antigravity access for six months, with applications open and selection required.
Why it matters for marketers: Argon's cheaper-than-rivals token rate has an expiry date, so teams budgeting agent workflows should plan for doubled costs and check which subscription tier unlocks the tools they use.
Google has released Gemini 4 Argon, its newest frontier model, to developers and paying consumers. Google says the model leads the Vals Index, which covers work in finance, coding, legal and tax, and that it beats rival models on a range of benchmarks. Those claims come from Google, and The Verge notes the company is comparing Argon with offerings from OpenAI and Anthropic.
Access depends on the plan. Ars Technica reports that Gemini app users need a Pro or Ultra subscription, while Antigravity and AI Studio are limited to Ultra. Google has not said whether lower tiers will get Argon in Antigravity, and The Verge says it is unclear if Plus or free users will get the model at all.
For teams that build on the API, the price matters most. Ars Technica reports the launch rate runs through November 30 and then doubles. The Verge adds that even the higher rate would sit below rivals per token, but cautions that models use different token counts for the same job, so real costs can differ.
Marketers should treat Google's internal results as claims, not proof for their own work. Google points to document synthesis, proposal decks built from messy project folders, and scheduled tasks in the Gemini app, which are closer to content operations than the coding examples. Test those on your own material before committing.
Watch for any change in tier access and for pricing once the introductory period ends. If you plan agent workflows on Argon, model the post-November costs now.
Sources
This brief is our own summary. Read the full reporting at the source.
- Google DeepMind: Gemini 4 Argon: Built to solve your hardest problems
- Google AI: Gemini 4 Argon: Built to solve your hardest problems
- The Verge: Google finally releases Gemini 4 Argon
- Ars Technica: Gemini 4 Argon begins rolling out in Gemini app and Antigravity, but only if you pay
- Search Engine Journal: Google Adds Gemini 4 Argon To AI Mode For Pro & Ultra Subscribers via @sejournal, @MattGSouthern
- The Decoder: Gemini 4 Argon is now available worldwide as Google's most capable frontier model
- Latent Space: [AINews] AI Engineer NYC: Gemini 4 Argon, Vals AI Benchmarks
- MarkTechPost: Google Releases Gemini 4 Argon API With 1M Token Output Limit
