Those who pay $250 a month for a Gemini Ultra subscription can now use Gemini 2.5 Deep Think. The so-called multi-agent AI model is designed to solve complex problems by using multiple thought processes simultaneously.
Starting this Friday, Google Deepmind has been rolling out Gemini 2.5 Deep Think to highly paid users. This model costs $250 per month via Ultra Subscription. This system works differently from traditional AI models by deploying multiple AI agents in parallel to a single question.
This approach requires much more computing power, but according to Google, it offers significantly better results. The company demonstrated this by winning a gold medal at the International Mathematics Olympiad with a variant of the model. That version takes hours to perform complex calculations rather than the seconds or minutes required for a traditional AI model.
Google claims that Gemini 2.5 Deep Think remains competitive. In the final exam of humanity, the model wins 34.8% without tools. Xai's Grok 4 scored 25.4%, while Openai's O3 remains at 20.3%. The difference is even greater in programming challenges. Gemini 2.5 Deep Think achieves 87.6% with LiveCodebench6. Grok 4 achieved 79% and O3 72%.
Multi-agent
Several AI players are experimenting with multi-agent systems. Elon Musk's Xai recently launched the Grok 4 Heavy based on this approach. It is also said that Openai uses a multi-agent model that suspects GPT-5. Human Research Agents are also working with multiple AI components to provide thorough research reports. We are currently limiting these systems to expensive subscriptions. This is not surprising. A single multi-agent task is essentially the sum of tasks that run in parallel.
Google plans to run further testing via the Gemini API in the coming weeks. The company wants to discover how developers and businesses can use multi-agent systems for practical applications.
