Gemini 4 Argon has a context window of one million tokens, according to Google
Google states that Gemini 4 Argon has a context window of one million tokens. The Decoder also reports up to one million output tokens. Access remains restricted to selected cybersecurity defenders in Fairwind Program.
Google states that Gemini 4 Argon has a context window of one million tokens and supports long, multistep tasks. In addition to this input context, The Decoder reports a separate limit of up to one million output tokens. According to this source, the model accepts text, images, video and audio, but produces text only. According to Google, it also has increased resistance to indirect prompt injection attacks and improved mechanisms for detecting misuse.
Google introduced the model on 30 September and is making it available to selected cybersecurity defenders through Fairwind Program. According to Google, the model was trained for defensive cybersecurity and can autonomously find, verify and fix critical software vulnerabilities. Public access is not yet available. Access is set to expand after feedback has been evaluated; according to The Decoder, paying Gemini API customers and Google AI Ultra subscribers are expected to be among the first to gain access. No date has been announced.
The Decoder reports an introductory price of 2 USD per million input tokens and 10 USD per million output tokens. Subsequent standard rates are set to be 4 USD and 20 USD per million tokens, respectively; the source does not give a date for the transition. The discount on input tokens retrieved from the cache is set to be 95 %, corresponding to 0.10 USD per million tokens at the introductory rate and 0.20 USD at the standard rate.
According to Google, the model achieved 77.9 % in the DeepSWE v1.1 test, helped save 300 TiB of memory in data centers, and its agents converted over 800 000 lines of Zircon kernel code from C/C++ to Rust. According to The Decoder, an independent evaluation by Artificial Analysis gives the model 53 points in the Intelligence Index. In this testing, however, the model consumed an average of 62 000 output tokens per task, so a low per-token rate alone does not guarantee a low cost per completed task.
Why it matters
Selected security organizations are gaining access to a model designed for finding and fixing vulnerabilities and handling long-running code tasks. For other users, it is not yet an available work tool. The cost per completed task will be important in business evaluations: long outputs may increase consumption, while reusing cached inputs carries a significantly lower rate.
Release card
Gemini 4 Argon
- Context
- The model supports up to one million input tokens and up to one million output tokens.
- Inputs
- The model accepts text, images, video and audio. It outputs text only.
- Price
- The announced introductory price is 2 USD per million input tokens and 10 USD per million output tokens. Subsequent standard prices are set to be 4 USD per million input tokens and 20 USD per million output tokens. Cached input tokens receive a 95 % discount on the applicable input price.
- Availability
- The model is being made available to selected security partners through Fairwind Program and is used by internal teams at Google. Broader access is planned to follow later, initially for paying API customers and subscribers to Google AI Ultra; no date has been announced.
- DeepSWE v1.1 77,9 % According to Google, this result measures the ability to handle long-running software engineering tasks.
- AutomationBench 51,3 % According to Google, the score measures the execution of entire workflows across core business activities.
- LVBench 91,7 % According to Google, the result reflects the performance of the model in understanding long videos.
- CWE-bench v1 68 % According to Google, the result measures the ability of the model to fix security vulnerabilities.
- Artificial Analysis Intelligence Index 53 bodů At the High reasoning level, the model achieved the same score as GPT-6 Astra at the max setting and Claude Fable 5.1.
- According to Google, the model finds, verifies and fixes software vulnerabilities.
- The model helps with debugging and code migration.
- The model analyzes charts and the content of long videos.
- The model does not provide image, audio or video outputs.
In the Artificial Analysis Intelligence Index evaluation, the model scored 53 points, the same as GPT-6 Astra and Claude Fable 5.1. Claude Opus 5.5 scored 58 points and Claude Sonnet 5.5 scored 56 points.
The card summarizes information from the article and any dated corrections, with a link to the original source. It is not our assessment of the model. It does not yet have a dedicated editorial profile. Model selection and other announcements →
What was added since the original report
Verified updates
-
Gemini 4 Argon is now available worldwide to developers, businesses and consumers.; The model is available through the API in AI Studio and on the Gemini platform.; Google AI Ultra subscribers have gained access in Google Antigravity.; AI Ultra and Pro subscribers have gained access in the Gemini app and AI mode in Google Search.; Under the new tier structure, free users and Plus subscribers have access only to Flash models.
- Gemini 4 Argon is now available worldwide to developers, businesses and consumers.
- The model is available through the API in AI Studio and on the Gemini platform.
- Google AI Ultra subscribers have gained access in Google Antigravity.
- AI Ultra and Pro subscribers have gained access in the Gemini app and AI mode in Google Search.
- Under the new tier structure, free users and Plus subscribers have access only to Flash models.
-
Google is beginning a broader rollout of Gemini 4 Argon.; Fairwind Program is expanding to include additional trusted cyber defense experts.; According to Google, the model successfully handled 12 out of 12 simulated cyberattack scenarios, compared with five for the previous generation.; According to Google, Wiz used the model to uncover a critical flaw in hospital software that put hundreds of thousands of patient records at risk.
- Google is beginning a broader rollout of Gemini 4 Argon.
- Fairwind Program is expanding to include additional trusted cyber defense experts.
- According to Google, the model successfully handled 12 out of 12 simulated cyberattack scenarios, compared with five for the previous generation.
- According to Google, Wiz used the model to uncover a critical flaw in hospital software that put hundreds of thousands of patient records at risk.
-
Gemini 4 Argon is now globally available through the API in AI Studio and Gemini platform.; Google AI Ultra subscribers have gained access in Google Antigravity.; Google AI Ultra and Pro subscribers have gained access in Gemini app and AI Mode in Search.; According to Google, the model ranked first in the Vals Index benchmark.; Google reports an increase of more than 35 % in the number of submitted code changes created by agents during internal use.
- Gemini 4 Argon is now globally available through the API in AI Studio and Gemini platform.
- Google AI Ultra subscribers have gained access in Google Antigravity.
- Google AI Ultra and Pro subscribers have gained access in Gemini app and AI Mode in Search.
- According to Google, the model ranked first in the Vals Index benchmark.
- Google reports an increase of more than 35 % in the number of submitted code changes created by agents during internal use.
-
The rollout of the model in the Gemini app and the Antigravity tool has begun.; The new access is currently limited to paying users.; Lower paid tiers do not have full access.; According to Google, thousands of employees used the model internally.
- The rollout of the model in the Gemini app and the Antigravity tool has begun.
- The new access is currently limited to paying users.
- Lower paid tiers do not have full access.
- According to Google, thousands of employees used the model internally.
-
Gemini 4 Argon is now available to developers through the paid Gemini API service.; Access through Gemini API is available in Google AI Studio and Android Studio.; Google AI Ultra subscribers can try the model in AI Studio and Google Antigravity.
- Gemini 4 Argon is now available to developers through the paid Gemini API service.
- Access through Gemini API is available in Google AI Studio and Android Studio.
- Google AI Ultra subscribers can try the model in AI Studio and Google Antigravity.
-
Gemini 4 Argon is now available in the Gemini app.; Google AI Pro and Google AI Ultra subscribers have access.; The Google Antigravity interface is available directly in the Gemini app for macOS.; Assigned tasks can continue while the user is away from the computer.; According to the announcement, the Gemini app supports connections to more than 40 apps.
- Gemini 4 Argon is now available in the Gemini app.
- Google AI Pro and Google AI Ultra subscribers have access.
- The Google Antigravity interface is available directly in the Gemini app for macOS.
- Assigned tasks can continue while the user is away from the computer.
- According to the announcement, the Gemini app supports connections to more than 40 apps.
-
The model is gradually being made available to Pro and Ultra subscribers in the Gemini app and AI mode in Search.; Ultra subscribers are gaining access on the Antigravity platform.; Prices will double after the introductory period ends at the end of November.
- The model is gradually being made available to Pro and Ultra subscribers in the Gemini app and AI mode in Search.
- Ultra subscribers are gaining access on the Antigravity platform.
- Prices will double after the introductory period ends at the end of November.
-
The model has a context window of one million tokens, according to Google.
-
The model was introduced on 30 September.; The model supports long, multistep tasks.; The model works with text, images and video.; Google reports increased resistance to indirect prompt injection attacks.; Google reports improvements to internal activation mechanisms for detecting misuse.
- The model was introduced on 30 September.
- The model supports long, multistep tasks.
- The model works with text, images and video.
- Google reports increased resistance to indirect prompt injection attacks.
- Google reports improvements to internal activation mechanisms for detecting misuse.
-
The model was trained specifically for defensive cybersecurity.; According to Google, the model autonomously finds, verifies and fixes critical software vulnerabilities.; According to Google, the model analyzes the content of long videos and charts.
- The model was trained specifically for defensive cybersecurity.
- According to Google, the model autonomously finds, verifies and fixes critical software vulnerabilities.
- According to Google, the model analyzes the content of long videos and charts.
-
The model supports up to one million output tokens.; The standard price is set to be 4 USD per million input tokens later.; The standard price is set to be 20 USD per million output tokens later.; Cached inputs receive a 95 % discount.
- The model supports up to one million output tokens.
- The standard price is set to be 4 USD per million input tokens later.
- The standard price is set to be 20 USD per million output tokens later.
- Cached inputs receive a 95 % discount.
-
Google is participating in a voluntary US government process for access to models before their release.
-
Google reports a score of 77.9 % in the DeepSWE v1.1 benchmark.; According to Google, the model helped save 300 TiB of memory in data centers.; According to Google, the agents converted over 800 000 lines of Zircon kernel code from C/C++ to Rust.
- Google reports a score of 77.9 % in the DeepSWE v1.1 benchmark.
- According to Google, the model helped save 300 TiB of memory in data centers.
- According to Google, the agents converted over 800 000 lines of Zircon kernel code from C/C++ to Rust.
Two audiences, two different impacts
What this means
For individuals
A developer who is not among the selected participants in Fairwind Program cannot yet try Gemini 4 Argon or count on it as an available tool for everyday work. No date for broader access has been announced.
More practical updates →For a business
Selected security organizations can evaluate the model for the process of finding and fixing software vulnerabilities. When deciding whether to deploy it, they need to verify its success rate on their own tasks and the cost of completing them, because long outputs may outweigh the advantage of low per-token rates.
Risks and complianceCheck the original
Event sources
independently confirmed · 2 publishers, 1 independent. We count feeds from the same owner only once.