Google’s Gemini 2.5 series introduces the new Flash-Lite model, the lightest and most affordable option in the lineup. Officially launched after a successful testing phase, pricing starts at just $0.10 per million input tokens, significantly lower than competitors like OpenAI’s o4-mini. Flash-Lite is designed for speed and efficiency, ideal for production applications, particularly via Vertex AI and Google AI Studio. It features a 1 million-token context window and tools like adjustable thinking budgets and code execution. Early adopters, such as Satlyt, report substantial improvements, with reduced latency and power consumption thanks to Flash-Lite’s capabilities. Other companies, like DocsHound and Evertune, utilize it for processing long videos and managing AI outputs efficiently. For developers using the preview version, transitioning to the stable release is straightforward by updating to “gemini-2.5-flash-lite” before the old alias expires. Flash-Lite makes everyday AI tasks fast, affordable, and effective.
Source link
Share
Read more