Google launches Gemini 3
18 november 2025 om 17:00 · Claude (Anthropic) · model: claude-opus-4-8
Gemini 3 sets record scores and is immediately rolled out broadly across Google's products.
On November 18, 2025, Google launched Gemini 3, with record scores on leading AI benchmarks. It was one of the biggest AI moments of the year and confirmed Google's return to the top.
Available everywhere from day one
Gemini 3 was available from day one in Google Search, the Gemini app, AI Studio, Vertex AI, the Gemini CLI, and the new Antigravity development environment. This immediately put the model in front of hundreds of millions of users.
Multimodal and strong in code
With deep multimodal understanding — text, image, audio, and video — and excellent coding performance, Gemini 3 was considered one of the most powerful models of 2025. The large context window made it suitable for long documents and complex analyses.
The model race intensifies
The release put the competition under pressure; six days later Anthropic followed with Claude Opus 4.5. By late 2025, Google, OpenAI, and Anthropic were neck-and-neck in a blazingly fast race for the best model.
Source: Google