Google on Wednesday introduced Argon, a new high-end artificial intelligence model that the company positions as the flagship of its Gemini 4 generation. Following the announcement, Alphabet shares rose 1.5 percent in after-hours trading.
Performance and comparisons
A company spokesperson said Argon is Google’s best-performing model to date and was designed to handle complex tasks. Google compared Argon to OpenAI’s Astra and Anthropic’s Opus, saying the models are on par particularly in programming and cybersecurity performance tests. According to Google’s internal measurements, Argon outperforms Astra and Opus on several industry benchmarks, but it still lags the rivals on certain metrics — including two of four programming tests.
Access and testing
Google has not announced when Argon will be broadly available. For now, the model is being tested by select cybersecurity partners, and Google is participating in the Trump administration’s voluntary pre-release model access program.
Cancelled model and internal changes
At the same time as the Argon announcement, Google confirmed it has permanently retired the Gemini 3.5 Pro model, whose release Sundar Pichai had previously promised for June. During months of delays, the company’s DeepMind AI lab underwent a substantial reorganization: its founder and leader, Demis Hassabis, stepped back, and several leaders responsible for developing the Gemini models left the company.
Competitive position and messaging shift
Because of the delays, Google has fallen behind Anthropic and OpenAI, which have continued to release new flagship models. In July Sundar Pichai rejected the idea that Google was losing ground in the AI race, but the company’s public messaging has shifted in recent months away from emphasizing technological leadership toward highlighting relatively better cost positions versus competitors.
Why it matters
Argon’s introduction signals Google’s attempt to regain momentum in the AI race, especially in areas like programming and cybersecurity. The developments help explain recent leadership and strategic changes inside the company and may influence investor expectations, as reflected in the post-close share price increase.
Related events
The article notes that related topics will be discussed at the Portfolio Investment Day on October 21 and at the Portfolio AI & Digital Transformation conference on November 26.
Source: Reuters. An AI assistant contributed to preparing this article; the final content was edited and verified by a journalist.
