Google's new AI can now "think deeply": More than just answers – faster, smarter, and sounds more human than ever before – Image: Xpert.Digital
Gemini 2.5 makes Google the AI market leader: What this means for users
Google introduces new Gemini 2.5 model and expands access
Google has announced significant progress with its Gemini 2.5 model, substantially expanding access to its most advanced AI technology. These latest developments mark a major milestone in Google's AI strategy and position the company as a leader in artificial intelligence.
Suitable for:
General availability of Gemini 2.5 Flash and Pro
On June 17, 2025, Google released the stable version of Gemini 2.5 Flash and Pro for general use. These models are no longer in the testing phase and can now be confidently used in production applications. Developers and businesses can use the models through Google AI Studio and Vertex AI, and they are also available in the Gemini app.
General availability means that Google will support these models long-term and that they are suitable for scaled production applications. Companies like Spline, Rooms, Snap, and SmartBear have already been successfully working with the latest versions in recent weeks.
Introduction of Gemini 2.5 Flash-Lite
As the latest addition to the Gemini 2.5 family, Google has introduced Gemini 2.5 Flash-Lite, the most cost-effective and fastest model in the 2.5 series. Initially available as a preview, Flash-Lite is aimed at developers who need to perform high-volume, latency-sensitive tasks such as translations and classifications.
The new model offers significantly better quality than Gemini 2.0 Flash-Lite in programming, mathematics, science, logical reasoning, and multimodal benchmarks. At the same time, it costs only a fraction of the full-price models and offers lower latency than its predecessors.
Advanced thinking skills and deep thinking
A key feature of Gemini 2.5 models is their advanced "thinking" capabilities. These models are able to fully consider their thought processes before responding, resulting in improved performance and greater accuracy. Developers can control the level of thinking intensity of the model before generating a response by using "thinking budgets."
Google has also announced an experimental “Deep Think” mode for Gemini 2.5 Pro. This mode allows the model to pursue multiple lines of reasoning in parallel before arriving at an answer, which is particularly beneficial for complex mathematical and programming tasks. In tests, Deep Think achieved top results at the 2025 US Mathematical Olympiad, scoring 84% in the demanding MMMU benchmark.
New features and improvements
Native audio output and Live API
Gemini 2.5 gains native audio output capabilities, enabling more natural conversations. The enhanced Live API supports audiovisual input and allows for direct interaction with the AI. Users can control tone of voice, accent, and expression, for example, telling the model to read stories in a dramatic tone.
The new experimental features include:
- Affective Dialogue: The model recognizes emotions in the voice and responds accordingly.
- Proactive Audio: Automatic filtering of background conversations
- Text-to-Speech: Multi-speaker support in over 24 languages
Improved programming skills
Gemini 2.5 Pro leads the WebDev Arena rankings and demonstrates significant improvements in web development. The model achieves 63.8% in SWE-Bench Verified, the industry standard for agent-based code evaluations. It excels in building visually appealing web apps and agent-based code applications, as well as in code transformation and editing.
The VideoMME benchmark demonstrates impressive multimodal capabilities: Gemini 2.5 Pro achieves 84.8% compared to 75% for Gemini 1.5 Pro and 71.9% for GPT-4o. This capability makes it possible to create entire applications from video content.
Enhanced multimodality and context processing
Gemini 2.5 builds on the strengths of the Gemini models: native multimodality and a large context window. The model launches with a 1-million-token context window, with 2 million tokens to be available soon. It can understand large datasets and handle complex problems from various information sources, including text, audio, images, videos, and entire code repositories.
Availability and access
For developers
- Google AI Studio: Immediate availability for experiments
- Vertex AI: Available for businesses with advanced features
- Gemini API: Full integration with SDK support
For end users
- Gemini App: Available for Gemini Advanced users on desktop and mobile.
- Google Search: Specially adapted versions of Flash Lite and Flash
Education sector
Google is extending free access to the Google AI Pro plan for students in Brazil, Indonesia, Japan, and the UK until the 2026 final exams. In addition to AI support, the package includes 2 TB of storage and NotebookLM.
Suitable for:
Technical specifications and performance
Gemini 2.5 Pro leads the LMArena rankings by a significant margin, demonstrating strong government performance across various benchmarks. The model achieves 18.8% in “Humanity's Last Exam,” a dataset developed by hundreds of subject matter experts to capture the limits of human knowledge and logical reasoning.
The latest version of Gemini 2.5 Pro shows a 24-point Elo jump on LMArena and a 35-point Elo jump on WebDevArena. It continues to lead in challenging programming benchmarks like Aider Polyglot and demonstrates top performance in GPQA and other demanding mathematical and scientific assessments.
Google Gemini 2.5 Flash and Pro transform the AI landscape with stable versions
The release of the stable versions of Gemini 2.5 Flash and Pro, along with the preview of Flash Lite, marks a significant step in Google's AI development. With a combination of improved performance, expanded features, and broader access, Google positions itself as a leader in the field of artificial intelligence.
The continuous improvements and expanded availability demonstrate Google's commitment to making AI technology more accessible and powerful for developers, businesses, and end users. With its new thinking capabilities and enhanced multimodality, Gemini 2.5 sets new standards for the next generation of AI applications.
Suitable for:
Your global marketing and business development partner
☑️ Our business language is English or German
☑️ NEW: Correspondence in your national language!
Konrad Wolfenstein
I would be happy to serve you and my team as a personal advisor.
You can contact me by filling out the contact form or simply call me on +49 89 89 674 804 (Munich) . My email address is: wolfenstein ∂ xpert.digital
I'm looking forward to our joint project.