Google Unveils Gemini 2.5 Flash & Flash-Lite: Faster, Smarter, and More Efficient
Google has unveiled the latest additions to its Gemini series, the Gemini 2.5 Flash and Gemini 2.5 Flash-Lite. These updated models promise enhanced intelligence and improved performance.
The new Gemini Flash boasts a significant boost in speed, with community reports suggesting it's twice as fast and four times cheaper than its predecessor on browser-agent tasks. Meanwhile, Gemini Flash-Lite can handle up to 1 million token contexts, with adjustable 'thinking budgets'.
Both models exhibit enhanced agentic tool use and multi-pass reasoning. Gemini 2.5 Flash-Lite, in particular, has shown impressive throughput in tests, making it the fastest proprietary model in its class. Google advises developers to pin fixed strings for stable production use.
The updates bring several improvements. They enable clearer, step-by-step explanations, enhance image understanding, and offer better answer formatting using headings, lists, and tables. They also work more efficiently, reducing token usage and costs. Additionally, they offer improved multimodal abilities and better instruction following. New aliases are available for developers to test the latest builds.
Google's Gemini 2.5 Flash and Gemini 2.5 Flash-Lite updates offer significant improvements in speed, capacity, and functionality. With a two-week notice before retargeting an alias, developers can plan their transitions. Pricing for Flash-Lite starts at $0.10 per 1 million input tokens and $0.40 per 1 million output tokens.
Read also:
- Web3 social arcade extends Pixelverse's tap-to-earn feature beyond Telegram to Base and Farcaster platforms.
- Trump praises the robustness of US-UK relations during his visit with Starmer at Chequers, showcasing the strong bond between the two nations.
- Navigating the Path to Tech Product Success: Expert Insights from Delasport, a Trailblazer in the Tech Industry
- German Startup Fernride Aims to Revolutionize Logistics with Autonomous Vehicles
