Google Launches Gemini 3

Read the original article →

Google Launches Gemini 3

Performance Leadership

Google launched Gemini 3, its latest flagship model. Gemini 3 outperforms Gemini 2.5 Pro on every major benchmark evaluated. The model achieved the top score on the LMArena Leaderboard with 1501 Elo, a rating system that compares large language models on user preferences.

Benchmark Results

Gemini 3 demonstrates improvements across code generation, reasoning, mathematics, and language understanding tasks. The model shows strong performance on specialized benchmarks for long-context understanding and multi-modal capabilities. These improvements establish Gemini 3 as a top performer in the competitive LLM market.

Product Offerings

Google introduced new subscription tiers for Gemini 3 access. Google AI Pro and Google AI Ultra subscriptions offer different feature sets and usage limits. This tiered approach allows users to choose appropriate service levels for their needs.

Companion Tools

Google released Veo 3, an advanced video generation tool, alongside Gemini 3. Nano Banana 2, an improved image generation tool, was also released. These tools expand Google's AI product portfolio and provide users with a full set of generative AI capabilities.

Market Position

The Gemini 3 launch directly competes with OpenAI's GPT-5 and Claude 4. Google's position as a major cloud provider allows integration of Gemini 3 into Google Cloud services. The leaderboard leadership demonstrates Google's continued investment in large language model development.

References

This article was originally published at Times of AI. For the full piece, read the original article.

Discussion

  • Loading…

← Back to News