Google is rolling Gemini 4 Argon out to developers, enterprises and consumers, pairing its new frontier model with introductory API prices well below several rivals. The launch matters because Google is making a direct play for complex, multi-step work, not just another round of chatbot small talk.
Google DeepMind Watch analysis
What happened
The Hindustan Times reports that Argon is now available globally through the Gemini API and to Google AI subscribers in Antigravity, the Gemini app and AI Mode in Search. It says Google is charging $2 per million input tokens and $10 per million output tokens at launch, with prices later rising to $4 and $20 respectively. Cached input tokens are priced at 95% off the input rate.
Google says Argon scored 68.90% on the Vals Index, ahead of the other models compared in the report, and reached 99.7% on the GraphWalks long-context benchmark. Those are company-supplied comparisons reported by the paper, not a promise that Argon will lead on every real-world job.
Why it matters
Access and price make this more than a leaderboard announcement. Developers can try the model through an API, while Google AI subscribers can use it in several existing products. The introductory price could make testing complex workloads less costly, though the later price is twice as high per token on both input and output.
Google also says Argon helped its Chrome team separate core webpage handling from a legacy folder of browser features, a task the company says took weeks rather than months. That is an intriguing example of long-running coding work, but it remains an internal account rather than an independent evaluation.
Our read
The notable combination is broad availability, a sharp launch discount and a set of claims aimed at work that takes multiple steps. That gives developers something concrete to test. Treat the benchmark lead as a starting signal, not a verdict: your own tasks, reliability and the price after the introductory period will matter more than a tidy ranking.
What to watch
- Whether independent evaluations reproduce Google’s benchmark results.
- When the introductory API prices end, and whether real workloads justify the later rates.
- How Argon performs on sustained coding and other multi-step tasks outside Google’s examples.
Discussion spark: Would you trial Argon at its introductory price, or wait for independent results and the standard rates before building around it?
Sources and evidence
- Google takes aim in frontier AI race as Gemini 4 Argon leads in benchmarks | Business News (13 October 2026, 16:00 UTC)
not affiliated with, endorsed by, or operated by Google or Google DeepMind