Google has introduced its next flagship artificial intelligence model, opening a new round in the race to build the most capable AI systems.
Announced on Wednesday, the new "Argon" model anchors Google's Gemini 4 generation. The company says it is larger in size than its previous top-tier Pro models and describes it as its most performant model yet for complex workloads, claiming it is comparable to OpenAI's Astra and Anthropic's Opus on key coding and cybersecurity benchmarks.
A bid to catch up in the AI race
The launch comes as Google tries to close the gap with rivals Anthropic and OpenAI, both of which have pushed out frontier models in recent months. Rather than going straight to a public release, Google is initially giving the model to select cybersecurity partners — a sign of how sensitive the most capable models have become.
The company also said it is taking part in the Trump administration's voluntary process for pre-release model access, which gives government reviewers a look at powerful AI systems before they are broadly deployed. Google has not provided a timeline for when the general public will be able to use Argon.
The announcement follows a flurry of model launches in September, including OpenAI's GPT-6 Astra and new agentic products from several labs, as competition intensifies ahead of expected initial public offerings by the leading AI firms.
Sources: Google announces Gemini 4 flagship AI model after months of delays
Announced on Wednesday, the new "Argon" model anchors Google's Gemini 4 generation. The company says it is larger in size than its previous top-tier Pro models and describes it as its most performant model yet for complex workloads, claiming it is comparable to OpenAI's Astra and Anthropic's Opus on key coding and cybersecurity benchmarks.
A bid to catch up in the AI race
The launch comes as Google tries to close the gap with rivals Anthropic and OpenAI, both of which have pushed out frontier models in recent months. Rather than going straight to a public release, Google is initially giving the model to select cybersecurity partners — a sign of how sensitive the most capable models have become.
The company also said it is taking part in the Trump administration's voluntary process for pre-release model access, which gives government reviewers a look at powerful AI systems before they are broadly deployed. Google has not provided a timeline for when the general public will be able to use Argon.
The announcement follows a flurry of model launches in September, including OpenAI's GPT-6 Astra and new agentic products from several labs, as competition intensifies ahead of expected initial public offerings by the leading AI firms.
Sources: Google announces Gemini 4 flagship AI model after months of delays
