# TestingCatalog — Google unveils Gemini 4 Argon with SOTA score on DeepSWE

- Company: TestingCatalog (testingcatalog.com)
- Announced: 2026-09-30T20:33:39+00:00
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://www.testingcatalog.com/google-unveils-gemini-4-argon-with-sota-score-on-deepswe/
- Record: https://forck.live/items/15479-google-unveils-gemini-4-argon-with-sota-score-on-deepswe
- Subject: AI models / agents / unreleased features
- Models affected: Gemini 4 Argon, GPT-6 Astra
- Pricing: introductory price of $2 per million input tokens and $10 per million output tokens, with cached input tokens priced at a 95% discount to the input rate. After the introductory period, pricing will rise to $4 per million input tokens and $20 per million output tokens.

Google unveiled Gemini 4 Argon, a frontier model for deep reasoning in software engineering, enterprise knowledge work, and cybersecurity. Access is initially limited to trusted cyber defenders via the Fairwind Program, with wider access planned for paid API and Google AI Ultra subscribers. The model achieved a 77.9% score on DeepSWE v1.1, tied for first on CWE-bench v1 with 68%, and scored 51.3% on Zapier's AutomationBench.

## Evidence

Verbatim from https://www.testingcatalog.com/google-unveils-gemini-4-argon-with-sota-score-on-deepswe/:

> The model scored 77.9% on DeepSWE v1.1 for long-horizon software engineering, ranked first on the Vals Index and scored 51.3% on Zapier’s AutomationBench.

---

Record: https://forck.live/items/15479-google-unveils-gemini-4-argon-with-sota-score-on-deepswe
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md