Google Opens Gemini 4 Argon to Paying Customers, Tops Vals AI Index
Google has released Gemini 4 Argon to paying customers and general availability after restricting it since its September 30 unveiling to Google teams, the US government, vetted cyber defenders and early-access partners. The model leads Vals AI's professional work index at 68.90%, ahead of Claude Sonnet 5.5 at 67.04% and Claude Opus 5.5 at 66.97%, and ranks equal to GPT-6 Astra on the Artificial Analysis Intelligence Index at 60% of the Cost per Task. It features a 1 million token output context window and introductory API pricing of $2/$10 per 1M input/output tokens, rising to $4/$20 after November 30.
IllustrationEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
Google Opens Gemini 4 Argon to Paying Customers, Tops Vals AI Index
Google has made Gemini 4 Argon available to paying customers after limiting it since its September 30 unveiling to Google teams, the US government, vetted cyber defenders and early-access partners. The model leads Vals AI's index of professional work with a score of 68.90%, ahead of Claude Sonnet 5.5 at 67.04% and Claude Opus 5.5 at 66.97%. Argon features a 1 million token output context window, with introductory API pricing of $2/$10 per 1M input/output tokens doubling to $4 and $20 after November 30. It averaged under 47 minutes and $15.68 per test, versus 1 hour 18 minutes and $21.34 for Sonnet 5.5. Vals priced Argon at the $4 and $20 standard rate Opus 5.5 also charges, yet Opus cost $32.14 per test, with Argon's cost edge coming from using fewer than half Opus's input tokens.
Read sourceGoogle's Gemini 4 Argon Goes GA, Matching GPT-6 Astra at 60% Cost per Task
Google released Gemini 4 Argon to general availability, its first proprietary model above the Flash class in over seven months. According to Artificial Analysis, Argon ranks equal to GPT-6 Astra on the Artificial Analysis Intelligence Index while costing 60% of the Cost per Task, with a discounted Cost per Task of $1.99, roughly 40% below GPT-6 Astra (max) at comparable intelligence. Pricing is temporarily discounted to $2/$10 per 1M input/output tokens with a 95% cache-read discount until November 30, after which prices rise to $4/$20. The model shows substantial gains on agentic evaluations including AA-Briefcase, AutomationBench-AA, and Terminal Bench 4, an area where Google models have historically underperformed. It also records a 15% hallucination rate on AA-Omniscience, the lowest among models scoring 45+ on the Intelligence Index.
Read sourceGoogle DeepMind's Gemini 4 Argon Launches on OpenRouter With Top Benchmark Scores
Google DeepMind's Gemini 4 Argon is now live on OpenRouter, accessible via Google AI Studio and Google Cloud. According to the announcement, the model tops AutomationBench at 51.3% and LVBench at 91.7%, and ties for first place on CWE-bench v1 with 68%. Introductory pricing is set at $2 per million input tokens and $10 per million output tokens. The post directs users to try the model through a provided link.
Read sourceShow 2 older updatesHide older updates
Gemini 4 Argon Launch Expected Today as Google Lists Model Pricing
A post on X claims that Google's Gemini 4 Argon model is set to launch today, citing several signals: model IDs are already present in the app, and pricing for Gemini 4 Argon has appeared on Google's SKU Pricing page, which the post says typically precedes a launch. The listed discounted pricing is $2 per million input tokens and $10 per million output tokens, while normal pricing is $4 per million input and $20 per million output. The post also notes that Gemini team members have begun vague posting, and it speculates that Ultra users are likely to get access first while Pro users might have to wait. The author adds that if Google does not release Argon today, it's over.
Read sourceGoogle's Gemini 4 Argon May Launch Today as Pricing Page Lists Model
An X post reports that Google's Gemini 4 Argon model could roll out today, citing a hint from Logan and references to Argon in the latest Antigravity update and Antigravity CLI. The post also notes that Google's SKU Pricing page already lists Argon at $2/$10 per million input/output tokens at the discounted price and $4/$20 at the normal price, suggesting discounted pricing would be available at launch. The author frames the release as the end of nearly seven months of waiting for Google's next Pro model, following the Gemini 3.5 cancellation and extensive testing around Gemini 4 Argon, and suggests Google might be making a comeback in the frontier model race — contingent on actually shipping it today.