Google announced that its Gemini 2.5 Pro and Flash AI models have moved from preview to general availability. The company stated that developers can now build production applications with confidence using these stable versions.
Several organizations including Snap, SmartBear, Spline, and Rooms have already been using the latest versions in production for recent weeks, according to Google. The announcement was made by Tulsee Doshi, Senior Director of Product Management for the Gemini team.
New Flash-Lite model targets cost efficiency
Google also introduced Gemini 2.5 Flash-Lite as a preview version. The company describes it as their most cost-efficient and fastest 2.5 model yet. According to Google’s benchmarks, Flash-Lite shows higher quality than the previous 2.0 Flash-Lite across coding, math, science, reasoning, and multimodal tasks.
The new model is designed for high-volume, latency-sensitive tasks like translation and classification. Google states that Flash-Lite costs one-third of the price of 2.5 Flash for text, image, and video inputs, and less than one-sixth for output tokens.
All Gemini 2.5 models include adjustable thinking budgets, Google Search integration, code execution capabilities, multimodal input support, and a 1 million-token context length. The Flash-Lite preview is available through Google AI Studio and Vertex AI for developers. Custom versions of Flash and Flash-Lite are now active in Google’s AI overviews and AI Mode search features.
The stable versions of 2.5 Flash and Pro are also accessible through the Gemini app, though free users have limited access compared to paid subscribers.
Additional source: Ars Technica