Key points
- Google unveils its next frontier model
- Cyber defenders get access first
- Pricing undercuts OpenAI's newest model
Alphabet (GOOGL) unveiled Gemini 4 Argon on Wednesday, a model Google called "our next era of frontier intelligence." The first users are a group of trusted cyber defenders in Google's Fairwind Program, and Google gave no date for a wider release.
Argon's introductory token prices of $2 per million input tokens and $10 per million output tokens are each one-fifth of OpenAI's listed rates for GPT-6 Astra. Koray Kavukcuoglu, senior vice president of Google DeepMind and Google's chief AI architect, announced the model in a company blog post.
Alphabet shares closed up 0.9% at $344.15 on Wednesday and were at $349.45 in after-hours trading as of 4:51 p.m. ET.
Who gets Gemini 4 Argon first?
Google said it will release Argon to trusted defenders and its own teams "without cyber guardrails so they can leverage its full frontier-level cybersecurity defense capabilities." Google launched Fairwind on September 2 for government agencies, Google Cloud customers, and security partners, and the program has more than 650 partners. The company is also taking part in the U.S. government's voluntary process for pre-release model access. Google plans to expand access to paid API customers and Google AI Ultra subscribers but has not announced when.
Wiz is already using Argon through its Scan for Good program, which looks for high-risk exposures in critical public infrastructure for free. According to Google, the model found a critical vulnerability in healthcare software used by hospitals worldwide, "identifying a severe risk that previous frontier models had missed."
Earlier this month, Google confirmed that a Gemini model gained unauthorized access to three real companies during a May security test. Google's safeguards for Argon monitor the model's chain of thought and actions and "stop execution when necessary," the company said.
How does Argon compare on price and benchmarks?
After the introductory period, Argon will cost $4 per million input tokens and $20 per million output tokens. GPT-6 Astra lists at $10 and $50, the same list price as Anthropic's Claude Fable 5.1. Google said Argon supports up to 1 million output tokens, compared with 64,000 for previous Gemini models.
In a benchmark table Google published, Argon had the top score on 13 of the 19 results against GPT-6 Astra, Claude Fable 5.1, and Claude Opus 5.5, and tied GPT-6 Astra on one. All of the figures are Google's, and the company posted its evaluation methodology alongside them.
GPT-6 Astra led on FrontierSWE v2, Terminal-Bench Science 0.1, and OSWorld-2.0. Claude Opus 5.5 led on Terminal-bench 4.0 and PostTrainBench.
| Benchmark | Gemini 4 Argon | GPT-6 Astra | Claude Fable 5.1 | Claude Opus 5.5 |
|---|---|---|---|---|
| Vals Index | 68.9% | 63.1% | 65.8% | 67.0% |
| AutomationBench | 51.3% | 41.4% | 31.4% | 42.5% |
| Vals Finance Agent v2 | 65.4% | 53.5% | 58.9% | 58.6% |
| Harvey's Legal Agent Benchmark | 19.6% | 5.4% | 6.7% | 3.8% |
| DeepSWE v1.1 | 77.9% | 74.1% | 67.4% | 74.2% |
| FrontierSWE v2 | 55.0% | 65.5% | 56.3% | 62.3% |
| Vibe Code Bench | 91.9% | 89.6% | 90.3% | 90.3% |
| Terminal-bench 4.0 | 57.4% | 58.2% | 57.9% | 66.4% |
| PostTrainBench | 45.3% | 44.3% | 40.2% | 49.3% |
| Terminal-Bench Science 0.1 | 57.6% | 68.1% | 52.6% | 63.3% |
| LABBench 2 | 88.8% | 85.4% | 68.6% | 73.1% |
| RiemannBench | 76.0% | 72.0% | 65.6% | 69.6% |
| GraphWalks, up to 128K | 99.7% | 98.7% | 91.4% | 90.6% |
| GraphWalks, 256K to 1M | 84.2% | 71.8% | 65.0% | 66.8% |
| Agent's Last Exam | 39.5% | 34.2% | n/a | 38.2% |
| OSWorld-2.0 | 69.2% | 72.6% | n/a | n/a |
| Chartography | 71.6% | 71.0% | 46.2% | 66.3% |
| LVBench | 91.7% | 87.5% | 79.7% | 83.7% |
| CWE-bench v1 | 68.0% | 68.0% | 58.0% | 67.0% |
A separate Google chart, citing Gray Swan's indirect prompt-injection benchmark, reports attack success rates at 15 attempts: 0.7% for Argon, 1.0% for Claude Opus 5.5, and 8.5% for GPT-6 Astra. The chart does not specify how many test cases those rates cover.
Google had promised Gemini 3.5 Pro for June and spent the summer releasing smaller Flash models instead. Bloomberg reported, citing people with direct access to the project, that Argon does well on benchmarks but less well when staff put it to work, and that it struggles with some coding tasks.
Google said thousands of its employees already use Argon. A team of Argon agents found memory optimizations that freed more than 300 TiB across Google's data centers once rolled out, with an estimated 500 TiB to 1 PiB in total savings, the company said. "Argon is fundamentally changing the way we work and build at Google," Kavukcuoglu wrote.
