Google’s Gemini 4 Argon Closes the Pricing Triangle. The Benchmark Lead Is the Real Story. – Forkast

Analysis

With identical $2/$10 pricing to OpenAI’s GPT-6.1 Sol and a DeepSWE score that leads the frontier, Google enters the model race as the fourth major lab to ship at the commodity floor – while Anthropic sits isolated behind a supply chain risk classification.

The Pricing Triangle Closes

Google DeepMind chose the last day of September to do something that makes the competitive landscape of frontier AI look almost mathematical. On September 30, SVP Koray Kavukcuoglu announced Gemini 4 Argon at introductory pricing of $2 per million input tokens and $10 per million output tokens, with cached input at 95% off. Twenty-four hours earlier, OpenAI had launched GPT-6.1 Sol at exactly the same rates.

The match is not a coincidence. It is the commodity floor arriving in the frontier model tier. OpenAI halved its own pricing from GPT-6 Astra to get here. Google, entering the race for the first time with a model capable of leading on benchmarks, chose to meet that price rather than undercut or exceed it. The introductory rates will eventually double to $4/$20 – but for now, the signal is clear: the two largest US labs have converged on the same economics for near-frontier intelligence.

Anthropic’s S-1 prospectus, filed two days earlier with $518 billion in compute commitments, makes the contrast sharper. Anthropic has not launched a $2/$10 competitor. Its strategy remains enterprise-led, anchored by Claude Code and quarterly revenue of $11.6 billion that now surpasses OpenAI’s. The pricing triangle – OpenAI at $2/$10, Google at $2/$10, Anthropic enterprise-only – represents a genuine bifurcation in business models, not merely a difference in marketing.

The Benchmark That Matters

Google is positioning Argon as the new frontier leader, and the number it is leading with is DeepSWE v1.1: 77.9%. That score, if it holds under independent scrutiny, edges out Claude Opus 5.5 at 74.2% and GPT-6 Astra at 74.1%. DeepSWE measures real-world long-horizon software engineering tasks – not synthetic benchmarks, but the kind of multi-step coding work that enterprise buyers actually need.

The caveat is significant. These are vendor-reported figures. Google has not published independent verification, and the industry’s track record on self-reported benchmarks is uneven. What is verifiable is the model’s technical specifications: a 2 million token context window and a 1 million output token limit, up from 64,000 in prior Gemini models. That output ceiling is industry-leading and designed for the kind of sustained, complex reasoning that enterprise workflows demand.

Google reports that internal use of Argon has already produced a 40% improvement in quantum algorithmic optimization, freed over 300 TiB of memory across its data centers, and is driving large-scale codebase migrations to Rust – including an 800,000-line rewrite of the Fuchsia Zircon kernel. On the Vals Index, which measures economic impact across finance, coding, legal, and tax work weighted by GDP contribution, Argon ranks first. On CWE-bench v1, which evaluates vulnerability remediation, it ties for first at 68%.

The Fairwind Irony

Argon’s distribution strategy is where the story gets structurally interesting. Google has gated the model behind its Fairwind Program, providing access to over 650 trusted cyber defenders – government agencies, critical infrastructure operators in healthcare, telecom, energy, and finance, and security vendors including CrowdStrike, Palo Alto Networks, and Wiz. Defensive use only. No offensive or dual-use authorization. Wiz is already using Argon through its Scan for Good initiative to find vulnerabilities in healthcare software used by hospitals worldwide.

This creates a sharp contrast with Anthropic. On September 25, the D.C. Circuit classified Anthropic as a supply chain risk under FASCSSA Section 4713, turning the company’s safety restrictions into a national security liability. The ruling effectively reclassified Anthropic’s refusal to allow its models to be used for autonomous weapons or domestic mass surveillance as a risk to national security – the very restrictions that Google is now positioning as a competitive advantage through Fairwind.

Google is shipping frontier cyber capabilities to the same defense ecosystem that Anthropic is now legally constrained from serving. The company that restricted itself is being punished for it; the company that gated access to trusted defenders is being rewarded. Whether that dynamic holds depends on how the FTC probe, announced the same day as the Argon launch, evaluates the safety claims of all frontier labs.

Four Labs, One Price Point

The Argon launch completes a month that has seen all four major US frontier labs deploy autonomous agents or next-generation models at or near the $2/$10 price point. OpenAI launched Dots and the $500-per-month Pro 500 tier on September 29. Meta’s Muse has been in consumer beta since September 8. SpaceXAI’s GrokBot shipped August 11. Apple’s Siri AI launched September 14 as personal-only, without enterprise management.

Google’s entry with Argon is different from the others in one structural respect: it is the first to match OpenAI’s pricing exactly while claiming benchmark leadership. OpenAI compressed its own margins to get to $2/$10; Google arrived at the same number on day one. The question is whether that convergence reflects a shared bet on volume-and-platform economics or simply the recognition that any higher price point is now untenable in a market where open-weight models from Chinese labs account for approximately 61% of top-model token traffic on OpenRouter, averaging $0.83 per million tokens – 7.3 times cheaper than proprietary alternatives.

What Remains Unresolved

Three uncertainties define the next phase. First, the benchmark gap: Argon’s DeepSWE 77.9% leads the frontier, but no independent lab has verified the number. The industry needs third-party evaluation before the benchmark leadership claim can be treated as structural rather than marketing. Second, the pricing floor: both Google and OpenAI plan to double their introductory rates. Whether the market will accept $4/$20 when open-weight alternatives cost a fraction remains an open question. Third, the regulatory environment: the FTC probe and the D.C. Circuit ruling create uncertainty for every frontier lab, but the asymmetry is real. Anthropic faces a supply chain risk classification while Google ships frontier cyber to the same defense customers Anthropic is restricted from serving.

The pricing triangle is closed. The benchmark race is Google’s to defend. The regulatory chessboard is still being set up.

Note: All benchmark figures cited are vendor-reported and have not been independently verified. Gemini 4 Argon introductory pricing ($2/$10) will increase to $4/$20 after the introductory period. The Fairwind Program launched September 2, 2026, for Gemini 3.8 Flash Cyber; Argon access rolled out September 30. Anthropic S-1 figures are from the prospectus as reported by Reuters – the filing has not yet appeared on SEC EDGAR. The FTC probe was confirmed by a senior FTC official to Reuters.