Google’s latest AI model Gemini 2.5 Pro sets new standards in artificial intelligence with advanced thinking capabilities and a context window of one million tokens.
The race for the most advanced AI technology reaches a new level with the release of Gemini 2.5 Pro. Google has unveiled its most powerful model yet, designed as a “thinking model” capable of thinking through its reasoning before generating an answer. This approach leads to significantly improved performance in various tasks.
Multimodal competence is another core element of the new model. Gemini 2.5 Pro can work seamlessly with text, code, audio, images and video, understanding and processing these different forms of input. This versatility makes the model suitable for a wide range of applications.

Extended context window and coding capabilities
A particularly impressive advance is the extended context window of one million tokens, which will soon be extended to two million tokens. This capacity enables the model to process and understand enormous amounts of information in a single interaction. Compared to previous models, Gemini 2.5 Pro can therefore capture more complex contexts and hold longer conversations.
Google also reports significant improvements in the model’s coding capabilities. Gemini 2.5 Pro shows advanced skills in creating web applications, editing code and developing software agents. These advances could significantly increase the productivity of developers and open up new possibilities for automated software development.
Ads
The most important facts about the update:
- Improved thought processes: The model can think through its thoughts before generating an answer
- Multimodal processing of text, code, audio, images and video
- Extended context window of 1 million tokens (soon 2 million)
- Leading position in benchmarks for math (AIME 2025) and science (GPQA diamond)
- Top position on the LMArena ranking for human preferences
- Available to Gemini Advanced users and via Google AI Studio
- Integration into Vertex AI in the coming weeks
The release of Gemini 2.5 Pro comes in an environment of intense competition, shortly after the release of OpenAI’s GPT-4.5 and Anthropic’s Claude 3.7 Sonnet. This evolution underscores Google’s ambition to maintain its position as a leader in AI development.
Source: Google Blog