Google Releases Two Versions of Gemini 2.5 Deep Think

Google made a version of its Gemini 2.5 Deep Think artificial intelligence model available Friday (Aug. 1) in the Gemini app to Google AI Ultra subscribers.

This model is a “significant improvement” over what Google announced in May at its Google I/O conference and is a variation of the one that achieved the gold-medal standard at this year’s International Math Olympiad, the company said in a Friday blog post.

The model released Friday to Google AI Ultra subscribers is designed for day-to-day use, reaching bronze-level performance on the 2025 IMO benchmark while operating faster, according to the post.

Deep Think uses parallel thinking techniques, meaning it generates several ideas, considers them simultaneously and then arrives at the best answer, the post said.

This model is likely to be used for iterative development and design, such as in web development; scientific and mathematical discovery; and algorithmic development and code, where it excels at tough coding problems, per the post.

“Deep Think can help people tackle problems that require creativity, strategic planning and making improvements step-by-step…,” the post said.

Google also said Friday that it released the version of Google 2.5 Deep Think that achieved the gold-medal standard with a small group of mathematicians and academics.

“We look forward to hearing how it could enhance their research and inquiry, and we’ll use their feedback as we continue to improve this offering,” the company said in the post.

Deep Think mode was announced at Google I/O in May along with other Gemini AI models. The lineup included Gemini 2.5 Pro; Updated Gemini 2.5 Flash, described as a faster, more efficient model; Deep Think mode, described as a high-performance version of Gemini 2.5 Pro for complex reasoning; and Gemini SDK, which provides compatibility with Model Context Protocol (MCP) to let agents use a variety of tools.

Google parent company Alphabet said in July that it raised its projected capital expenditures to $85 billion in 2025 to support its AI infrastructure.

The company spent $52.5 billion in 2024 and previously said it would spend $75 billion in 2025.

“Our AI infrastructure investments are crucial to meeting the growth and demand from cloud customers,” Alphabet and Google CEO Sundar Pichai said July 23 during the company’s second-quarter earnings call.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

The post Google Releases Two Versions of Gemini 2.5 Deep Think appeared first on PYMNTS.com.