On Tuesday, Alphabet Inc.’s (NASDAQ:GOOG) (NASDAQ:GOOGL) Google unveiled Gemini 2.5 Pro, the first in a new family of advanced AI reasoning models designed to pause and think before answering complex questions.
What Happened: Taking to X, formerly Twitter, Google CEO Sundar Pichai shared the development saying, “Gemini 2.5 is here, and it’s our most intelligent AI model ever.”
Gemini 2.5 Pro is a multimodal model optimized for agentic reasoning, code generation, and visually rich web development. It is now available through Google AI Studio and the Gemini app for users subscribed to Gemini Advanced.
Pichai also shared an example stating that users can simply prompt the AI to create a basic video game—such as the well-known Chrome Dino game — and it will generate the fully executable code from just a single-line instruction.
On Aider Polyglot, a benchmark for code editing, Gemini 2.5 Pro scored 68.6%, ahead of top models from OpenAI, Anthropic, and DeepSeek.
It also earned a 63.8% on SWE-bench Verified, a software development benchmark, surpassing OpenAI’s o3-mini and DeepSeek’s R1, though it fell short of Anthropic’s Claude 3.7 Sonnet.
On the challenging multimodal benchmark Humanity’s Last Exam, Gemini 2.5 Pro scored 18.8%, outperforming most flagship models.
The model launches with a 1 million token context window—roughly 750,000 words—and Google says support for a 2 million token window is coming soon.
Subscribe to the Benzinga Tech Trends newsletter to get all the latest tech developments delivered to your inbox.
Why It’s Important: Google introduced Gemini in December 2023, positioning it as a direct competitor to OpenAI’s advancements in artificial intelligence.
However, by early 2024, the company encountered setbacks with the model, particularly regarding inaccurate image generation, which led to a temporary suspension of that feature.
The controversy surrounding Gemini’s errors prompted Google CEO Sundar Pichai to address the issue publicly, calling the inaccuracies “completely unacceptable.”
Despite these obstacles, industry experts remain confident in Alphabet’s competitive standing in the AI landscape, pointing to its strong presence in cloud computing, search, and YouTube as key advantages.
Earlier this month, it was reported that Chinese smartphone brands have started leveraging Google’s AI technology to compete with global tech giants.
Price Action: As of this writing, Alphabet’s Class A shares were up 0.19% in after-hours trading, while Class C shares rose 0.24%. Earlier on Tuesday, Class A shares closed 1.72% higher at $170.56, while Class C shares gained 1.68%, ending at $172.79, according to Benzinga Pro data.
According to Benzinga’s Edge Rankings, Google holds a 72.35% growth rating. Curious how it stacks up against other companies? Click here for the full picture.
Image via Shutterstock
Check out more of Benzinga’s Consumer Tech coverage by following this link.
Read Next:
Disclaimer: This content was partially produced with the help of Benzinga Neuro and was reviewed and published by Benzinga editors.
