Google has introduced the Gemini 2.X model family, featuring Gemini 2.5 Pro and Gemini 2.5 Flash alongside the earlier Gemini 2.0 Flash and Flash-Lite models. This generation aims to power a new era of agentic systems by combining native multimodality, long context support exceeding one million tokens, and tool-use capabilities.
- Gemini 2.5 Pro is a thinking model that achieves state-of-the-art performance on frontier coding and reasoning benchmarks, supports processing up to three hours of video, and exhibits emergent multimodal coding abilities.
- Gemini 2.5 Flash offers excellent reasoning with a controllable thinking budget, allowing users to manage the tradeoff between quality, cost, and latency.
- Gemini 2.0 Flash provides high performance at low latency for everyday tasks, while Gemini 2.0 Flash-Lite is optimized for fast, cost-efficient at-scale usage.
The models cover the full Pareto frontier of capability versus cost, enabling users to explore complex agentic problem-solving across text, audio, images, video, and code repositories.