Google has officially launched the new Gemini 4 Argon, which is reportedly the first model in the Gemini 4 generation. The tech giant stated that teh new model has been designed to enable deep reasoning with complex and long-horizon workflows. This update has been specifically focused on software engineers, financial and legal work, along with cybersecurity defence systems.
The launch further follows an intense competition period in the booming AI industry, as Google faces an increase in capable models from leading companies like OpenAI and Anthropic.
Google stated that Argon is already in use by many, and it internally handles engineering and other tasks.
Gemini 4 Argon and its output capacity: 1 million token output limit
Google said that the new model can generate up to 1 million tokens in a single output, which is claimed to be one of the biggest upgrades in Gemini 4 Argon.
This means it gives enough room to work on longer and more complicated tasks, including technical work.
The tech giant has further highlighted the improvements in coding, multimodal understanding and reasoning with the new model. The third-party benchmark data, on the other hand, is said to list Argon across different domains like coding, finance, cybersecurity and other professional tasks.
Google is its own benchmark. Comparisons show that Argon is performing strongly across several tests, although the benchmark results can vary depending on the test methodology and competing model versions.
Gemini 4 Argon: Price and availability
Gemini 4 Argon is not available for general consumers as of now. But Google has confirmed that the broader access will be out soon (timeline unspecified). It will start with the paid API customers and Google AI Ultra subscribers, but so far, it has not announced any fixed timeline for when the model will come to everyone.
.The introductory price for API is around USD 2 per million input tokens and USD 10 per million output tokens. It will reportedly give cached input tokens which will further receive a 95 per cent discount.
After the introductory period, price is expected to go up to USD 4 per million input tokens and USD 20 per million output tokens.
Gemini 4 Argon focuses on cybersecurity
Gemini 4 Argon further focuses on cybersecurity, where Google says that it is capable of identifying, validating and addressing software vulnerabilities, which enables security teams to investigate threats more efficiently.
With these new capabilities in the newest model from Google, the company seems to be taking a phased approach to the rollout. Argon was initially made available to a group of trusted cyber defenders from Google's Fairwind programme only, and the tech giant has been collecting feedback and improving its safeguards.
ALSO READ:
Googlebook laptops with Gemini AI, OLED displays and 14-hour battery: Check details here
Google adds Gemini Voice features to Gmail, Docs and Keep: What you can do with it?
Gemini 3.7 Flash launched with Smarter AI Agents, complex coding at lower cost: What to expect
