Google's AI Model Gemini 2.5 Pro Preview: A Step Ahead in the AI Race
Google has released the latest version of its AI model, Gemini 2.5 Pro Preview (I/O Edition), earlier than expected. Originally planned for the Google I/O conference in May, the company responded to high demand, particularly from the developer community, with the advanced release. The new version promises significant improvements, especially in the area of web development and coding support.
Improved Web Development and Coding Features
Gemini 2.5 Pro Preview scores with optimized functionality for creating interactive web applications. In the "Webdev Arena Leaderboard," a benchmark test to evaluate the aesthetics and functionality of generated web apps, the model achieved a plus of 147 Elo points compared to its predecessor. The Elo system, known from chess, measures human preference for the results of different models in the AI context. The significant increase indicates a clear improvement in web development capabilities.
In addition, Gemini 2.5 Pro Preview offers enhanced support for complex development tasks. Code transformation, editing, and agent-based workflows are designed to be more efficient. The execution of function calls, a point of criticism in earlier versions, has also been improved according to Google.
Multimodal Capabilities and Contextual Understanding
Besides the coding improvements, Google emphasizes the advancements in multimodality. Gemini 2.5 Pro Preview can process not only text, but also images and videos. In the "VideoMME" test, the model achieved a score of 84.8 percent, indicating improved video understanding. This multimodal understanding enables a wider range of applications and opens up new possibilities in areas such as video analysis and editing.
Availability and Pricing
The updated version is accessible via the Gemini API, Google AI Studio, and Vertex AI. Users of the Gemini app also benefit from the new features. The price remains unchanged compared to the previous version of Gemini 2.5 Pro.
Strategic Positioning in the AI Competition
With the release of Gemini 2.5 Pro Preview, Google underscores its ambitions in the AI race. The focus on web development, combined with multimodality and improved contextual understanding, positions the model in competition with offerings from OpenAI, Anthropic, and xAI. The early release gives developers faster access to the new features and could give Google an advantage in the dynamic AI market.
Outlook
The development of AI models is progressing rapidly. Gemini 2.5 Pro Preview is another step towards more powerful and versatile AI systems. It remains to be seen how the new features will perform in practice and what innovations the future of AI development holds. However, the improved web development and multimodal strength of Gemini 2.5 Pro Preview suggest great potential for future applications.
Bibliography:
- t3n.de/news/gemini-2-5-pro-preview-das-kann-die-neue-version-von-googles-ki-modell-1686164/
- www.facebook.com/t3nMagazin/posts/eigentlich-wollte-google-die-neue-version-seines-ki-modells-gemini-25-pro-previe/1118295130335638/
- t3n.de/tag/kuenstliche-intelligenz/
- t3n.de/
- the-decoder.de/google-verbessert-sein-ki-modell-gemini-pro-2-5-fuers-coding/
- t3n.de/news/google-gibt-gemini-2-5-pro-fuer-alle-frei-zumindest-ein-bisschen-1680706/
- www.facebook.com/t3nMagazin/posts/google-bringt-eine-duolingo-konkurrenz-an-den-start-drei-ki-tools-die-aber-noch-/1112879090877242/
- t3n.de/tag/software-entwicklung/
- t3n.de/news/
.png)
