Google has fully released Gemini 4 Argon, its long-awaited frontier AI model, putting it in front of consumers and developers months after Google first trailed its rivals in the current generation of foundation models. According to The Verge, which reviewed the launch, Google claims Argon can compete with the best models the industry has to offer — and the company is backing that claim with pricing designed to undercut OpenAI and Anthropic by a wide margin.
The launch ends a waiting period that began when Google first teased Argon a few weeks ago with a limited release to a select group of trusted cyber defenders, a pre-deployment testing step that has become standard practice for frontier model launches. Now the general rollout is underway, with Google positioning Argon as its answer to GPT-6 Astra and Anthropic's Claude Fable lineup.
What Google Claims Argon Can Do
Google says Argon delivers what it calls frontier reasoning and improved performance across complex workflows, while cutting down on the hedging, repetition and clichés that continue to plague AI chatbots. According to The Verge's report, Google claims Argon outperforms rival models including OpenAI's GPT-6 Astra, Claude Fable 5.1 and Claude Opus 5.5 across a range of benchmarks spanning coding, cybersecurity, science, finance and law.
Google's own blog series around the launch emphasizes coding, math and robotics as the model's core strengths, and CNET's coverage of the release echoes that framing. The robotics angle was prominent in Google's announcement materials, titled "Gemini 4 Argon for Robotics."
The company is also using Argon internally. Google says the model is already finding and patching cybersecurity vulnerabilities, halved the computing power needed to run YouTube's comments section, and is being used to debug the drone physics simulator for Wing, Alphabet's drone delivery subsidiary.
Pricing Designed to Undercut the Competition
The most aggressive part of the launch is the price. Argon is launching with an introductory rate of $2 per million input tokens and $10 per million output tokens — which The Verge calculates at roughly a fifth the cost of OpenAI's GPT-6 Astra and Anthropic's Fable 5 models.
There is a catch: those introductory prices double at the end of November, when the promotional period ends. Even then, Google's pricing remains significantly cheaper than its frontier rivals on a per-token basis, though as The Verge notes, cheaper tokens do not necessarily translate into lower costs in practice, since less capable models can require more attempts to complete a task.
Where Argon Is Available Today
The rollout is starting at the top of Google's customer tiers. Argon is available now in the Gemini app and in AI Mode in Search for Google AI Pro and AI Ultra subscribers, and to Ultra subscribers on Antigravity, Google's agentic software development platform. Google has not said when — or whether — Argon will reach Plus subscribers or free users, a decision that will shape how widely the model's real-world performance is tested.
The Distribution Question Nobody Can Answer Yet
The tiered rollout leaves open the biggest strategic question of the launch: Argon's reach. Google's historic advantage has always been distribution — Gemini sits inside Search, Android, Workspace and Chrome, touching billions of users. Keeping Argon restricted to paid Pro and Ultra tiers concentrates the model among the users most likely to stress-test it, but it also delays the broad real-world evaluation that competitors' models have already accumulated. OpenAI and Anthropic both exposed their frontier models to free and mid-tier users early in their lifecycles.
For developers, the pricing math is more immediately clear. On the API side, Argon's introductory rates apply regardless of subscription tier, and Antigravity gives Ultra-tier developers agentic workflows built around the model. If Argon performs as claimed at a fifth of the price, the economics of production AI workloads — coding agents, document analysis, customer-facing assistants — shift meaningfully toward Google's platform.
India Today reported that the model is rolling out globally and is usable on phones from launch day, in line with how Google now ships flagship models across its app ecosystem.
A Candidacy Statement After a Year on the Sidelines
Argon is Google's first major frontier release since Gemini 3 arrived in late 2025. In the months between, OpenAI and Anthropic shipped successive models that reviewers and customers widely treated as the industry's best, and internal leadership shakeups at Google raised questions about the company's ability to respond. The Verge described Google as having spent those months "sitting on the sidelines of the frontier AI race."
The timing of the release is pointed. Argon landed just one day after Meta announced Muse Wonder Full, its own first full frontier model, making this a week in which three of the four major US AI labs have now staked their claims in the current model generation.
For Google, the argument Argon makes is less about any single benchmark win and more about the combination of frontier performance and aggressive pricing. If the benchmark claims hold up in independent testing, Google will have a credible frontier model at a price point that pressures OpenAI and Anthropic's margins. If they do not, the company risks another cycle of playing catch-up in perception, regardless of what its internal evaluations say.
Independent benchmarks and developer feedback over the coming weeks will show whether Argon's arrival reshuffles the frontier or simply narrows the gap. For ongoing coverage of the model race and the industry around it, keep up with our breaking AI news coverage.
