Google has added a new addition to its Gemini AI line with the release of Gemini 2.5 Flash-Lite, a model that promises high-level performance at low cost and unparalleled speed.
The new model was announced together with the mass roll-out of the Gemini 2.5 Flash and Pro models but as the most cost-efficient option in the Gemini 2.5 portfolio.
From Google, the 2.5 Flash-Lite model is designed to support high-volume, low-latency workloads such as language translation, data classification, and reasoning, but using less computational resources. The corporation asserts the new model offers improved accuracy and output quality in domains such as coding, math, scientific queries, reasoning, and multimodal benchmarks compared to its successor, the 2.0 Flash-Lite.
It performs exceptionally well on high-volume, low-latency tasks such as translation and classification with less latency than 2.0 Flash-Lite and 2.0 Flash on a large sample of prompts," Google added.
Lightweight, But Feature-Rich
In spite of its small size, Gemini 2.5 Flash-Lite is loaded with potent features. It has multimodal inputs, a 1 million-token context window, and computes nicely with services like Google Search and code execution environments. Developers are also given the flexibility to modify computational load according to their budget, so it is perfect for projects that require performance and affordability.
The model is currently in preview and accessible from Google AI Studio and Vertex AI. Furthermore, Google has integrated tailored copies of Flash-Lite and Flash into products such as Google Search, extending their reach to general users beyond developers.
Gemini 2.5 Flash and Pro: Now Generally Available
As part of a major update, Google also revealed that the Gemini 2.5 Flash and Gemini 2.5 Pro models have emerged from limited preview and are now available for all developers and enterprises. The models were earlier confined to early adopters for pre-production testing.
A number of companies—such as Snap, SmartBear, and Spline (a platform for creative tools)—have already begun leveraging these models within their production pipelines, and are seeing positive results.
With the 2.5 Flash and Pro models now completely available, developers can now implement them with confidence in production-level AI applications, with the added advantage of more stable performance and wider support.
Access to all Gemini 2.5 models, including the new Flash-Lite preview, is available through Google AI Studio, Vertex AI, and the Gemini app.
Read also| Lava Launches Storm Play and Storm Lite Smartphones in India, Priced Below ₹10,000
