Today, we release a stable version of Gemini 2.5 Flash-Lite, our fast and low costs ($ 0.10. Model rotates our set of 2.5 models ready for measured use.
Our most efficient and quick-speed model of 2,5
Gemini 2.5 Flash-Lite hit the balance between working and costs, without compromising quality, especially the sensitive latency services such as translation and division.
Here's what makes it outstanding:
- Best Class of Class: Gemini 2.5 Flash-Lite has the lowest 2,0 Flash-Lite Lite and 2.0 Flash in a broad range of arguments.
- Cost performance: Our lowest 2,5 model is currently low, the price in the End of $ 0.0 / 1m and ejecting of $ 0.40, allows you to carry large amounts of unpopular requests. We also reduce the Input prices of audio by 40% from the prevention of the launch.
- Smart and smaller: Displays the highest quality around over 2.0 flash-lite variety of benches, including codes, mathematics, science, consultation, and multiple multimodal understanding.
- Fully installed: When building a 2,5 Flash-Lite, you get access to a 1 million window, and traditional tools are like Google's search, code, and URL context.
Gemini 2.5 Flash-Lite in Action
From the introduction of 2,5 flash-Lite, we have already seen and successfully submit, here are some of our favorites:
- Satyt Creating a specified Computer platform is used to change satellite data is processed and how to use the actual in-orbit telemetry, independent tasks, and satellite-to-satellite Communication Parsing. The speed of 2,5 LITE has enabled 45% reduction in latency order Onboard sensitive diagnosis and a 30% depreciation of energy use Compare their models to Keen.
- Figure Using AI Creates Avatars Vatar Content and Gemini content and services for Gemini 2.5 Flash-Lite to change the video editing, Analyze and Expand the content, and Translate videos to more than 180 languages. This allows them to give worldwide, personally made to their users.
- Bill Curve the Demos of the product to be texts using Gemini 2.5 Flash-Lite to Use long videos and remove thousands of screens with low latency. This turns a footage to complete texts and training details for AI agents faster than traditional ways.
- An everco Help products Understand what they are represented in all AI models. Gemini 2.5 Flash-Lite is a part of the game to them, surprisingly accelerating the analysis and reported. Its quick performance allows to scan quickly and conform to large volume volumes to provide clients with Dynamic, timely understanding.
You can start using 2,5 flash-lite by specifying “gemini-2.5-flash-lite” in your code. If you use the preview version, you can switch to “Gemini-2.5-flash-lite” which is the same basic model. We plan to remove alialies first flash-Lite alias on August 25th.
Ready to start building? Try a stable version of Gemini 2.5 Flash-Lite now on Google Ai Studio Navertex AI.
