Gemini 4 Argon
Frontier intelligence for the hardest problems in code, knowledge work, and cyber defense.
Google's newest frontier model for coding and cyber defense; limited to vetted defenders for now, not public yet.
Gemini 4 Argon official announcement key art
Overview
What is Gemini 4 Argon?
Frontier intelligence for the hardest problems in code, knowledge work, and cyber defense.
Gemini 4 Argon is Google DeepMind's newest frontier AI model, announced September 30, 2026. Google describes it as its most powerful model yet, built for long, complex work across software engineering, enterprise knowledge work (legal, finance), and defensive cybersecurity — it can autonomously find, validate, and patch software vulnerabilities.
Important: Argon is not publicly available yet. Initial access is limited to vetted cybersecurity defenders through Google's Fairwind Program; paid API access and Google AI Ultra subscribers come next, with no announced date for general availability. This listing tracks the model itself — for the consumer Gemini app, see the separate Gemini entry.
What Google claims (official figures, not independently verified): 77.9% on DeepSWE v1.1 (long-horizon coding), joint-first 68% on CWE-bench v1 (vulnerability remediation), 51.3% on AutomationBench, top of the Vals Index, and 91.7% on LVBench long-video understanding. Standout spec: up to 1 million output tokens per response (up from 64K). Independent check: Artificial Analysis scores it 53 on its Intelligence Index (tying GPT-6 Astra and Claude Fable 5.1) with a 15% hallucination rate.
Pricing (announced): introductory $2 per million input tokens and $10 per million output tokens (cached input $0.10); rising to $4/$20 after the introductory period.
Platforms and languages
Gemini 4 Argon Availability
Platforms
Languages
Capabilities
Gemini 4 Argon Key Features
1M-token single outputs
generate up to one million tokens in one response for long refactors and reports
Defensive cybersecurity
autonomously find, validate, and patch software vulnerabilities
Long-horizon coding
built for multi-step software engineering across large codebases
Enterprise knowledge work
legal, finance, and document-heavy professional workflows
Long video understanding
chart analysis and detail extraction from long videos
Plans and access
Gemini 4 Argon Pricing
Paid
Common questions
Gemini 4 Argon FAQs
Can I use Gemini 4 Argon right now?
No. Access is currently limited to vetted cybersecurity defenders in Google's Fairwind Program. Paid API customers and Google AI Ultra subscribers are next in line, but Google has not announced a date for general availability.
How much will Gemini 4 Argon cost?
Google announced introductory pricing of $2 per million input tokens and $10 per million output tokens (cached input $0.10). After the introductory period, pricing rises to $4 input and $20 output per million tokens.
What is Gemini 4 Argon best at?
According to Google, its strengths are long-horizon software engineering, enterprise knowledge work such as legal and finance tasks, and defensive cybersecurity — finding, validating, and patching vulnerabilities.
How is this different from the Gemini app?
Argon is a frontier model (API), not the consumer Gemini app or chatbot. This is a separate listing for the model itself; the Gemini app has its own entry.
What are its headline specs?
Up to 1 million output tokens per response (up from 64K), and — per Google's official claims — 77.9% on DeepSWE v1.1 and joint-first 68% on CWE-bench v1 for vulnerability remediation.
Reviews
See what the community thinks and share your experience.
—
Based on 0 ratings
Rating distribution
Community reviews
No reviews
No written reviews yet.
