Gemini 2.5 Flash-Lite Review: Specs, Pricing & Best Uses

Gemini 2.5 Flash-Lite is the smallest, fastest, and cheapest model in Google's Gemini 2.5 family. Google previewed it on June 17, 2025 and made it generally available on July 22, 2025, pricing it at just $0.10 per million input tokens. It was built for high-volume jobs such as translation, classification, and quick summaries, where speed and cost matter more than depth.

This Gemini 2.5 Flash-Lite review covers what the model does, how it performed at launch, what it costs, how it compares with Gemini 2.5 Flash and the newer Gemini 3.1 Flash-Lite, and when it is the right pick. Gemini 2.5 Flash-Lite is available in Chat Smith on Android, iOS, and the web.

Gemini 2.5 Flash-Lite is the entry-level member of the family covered in our Gemini 2.5 review. It replaced Gemini 2.0 Flash-Lite and brought the 2.5 generation's improvements, including optional reasoning, to Google's lowest price tier.

It accepts text, images, audio, and video, supports a 1 million-token context window, and can write up to about 64,000 tokens per response. Thinking is off by default to keep answers fast, but developers can turn it on with a token budget when a request needs more care. Its knowledge runs to January 2025.

Flash-Lite also works with Google's built-in tools, including grounding with Google Search, code execution, and reading content from URLs, so even a lightweight model can pull in fresh information when it needs to.

Four strengths define Gemini 2.5 Flash-Lite.

At $0.10 per million input tokens and $0.40 per million output tokens, Flash-Lite costs about a third as much as Gemini 2.5 Flash for input and less than a sixth as much for output. That makes it practical for tasks that run thousands of times a day, like tagging content, routing support tickets, or translating short messages.

Google said Flash-Lite answered faster than Gemini 2.0 Flash and 2.0 Flash-Lite on a broad set of prompts. With thinking off, replies arrive almost instantly, which suits chat interfaces, autocomplete, and quick rewrites, the kind of jobs an AI summarizer does many times a day.

In Google's launch results, Gemini 2.5 Flash-Lite outscored 2.0 Flash-Lite across coding, math, science, reasoning, and multimodal tests. Turning thinking on lifted it further, to 66.7% on the GPQA Diamond science benchmark, so the same cheap model can handle the occasional harder question.

Despite its size, Flash-Lite keeps the family's 1 million-token window and can read images, audio, and video. That lets it transcribe recordings, describe screenshots, or summarize YouTube videos at a very low cost per request.

The table compares Gemini 2.5 Flash-Lite with its bigger sibling and with Gemini 3.1 Flash-Lite, its successor at the low-cost end of Google's lineup. Prices are Google's API rates.

Gemini 2.5 Flash-Lite is still the cheapest of the three, while Gemini 3.1 Flash-Lite closes much of the quality gap with Flash at a higher Lite-tier price. Our Gemini 2.5 Flash review and Gemini 3.1 Flash-Lite overview cover both in detail.

Gemini 2.5 Flash-Lite is available through the Gemini API in Google AI Studio and Vertex AI. Unlike Gemini 2.5 Flash, it charges the same output rate whether thinking is on or off:

In Chat Smith, you can use Gemini 2.5 Flash-Lite without an API key or Google AI subscription. Free accounts include limited use, and Chat Smith Pro raises usage limits for $9.99 per month or $39.99 per year.

Flash-Lite is not built for deep reasoning. Multi-step analysis, complex math, and nuanced writing are better handled by Gemini 2.5 Flash, Gemini 2.5 Pro, or a newer model, and short default answers can miss detail on vague prompts.

It is also no longer Google's newest Lite model. Gemini 3.1 Flash-Lite offers noticeably better quality, and our guide to Gemini models shows how the lineup has moved on.

Finally, its knowledge stops in January 2025, so recent events need search grounding, and its benchmark results come from Google's own testing. Check it on your real tasks before relying on it at scale.

Gemini 2.5 Flash-Lite is available in Chat Smith on Android, iOS, and the web. Pick it from the model list and start chatting, with no Google account or subscription needed.

Use Flash-Lite for quick replies, short summaries, and simple rewrites. When you need more depth, switch to Gemini 2.5 Flash or Gemini 3.5 Flash in the same chat, or try Gemini 3.1 Flash-Lite for a newer lightweight option.

Because Chat Smith also includes models from OpenAI, Anthropic, xAI, and DeepSeek, you can compare Flash-Lite's answers with other fast models and keep the one that holds up. For research-heavy questions, the built-in Deep Research feature runs advanced search and reasoning across sources.

Gemini 2.5 Flash-Lite delivers what it promises: very fast answers, the lowest price in the Gemini 2.5 family, a 1 million-token window, multimodal input, and optional thinking for the occasional harder question. For simple, high-volume work, it is hard to beat on cost.

For anything more demanding, step up to Flash or a newer model. You can try them side by side in Chat Smith, an AI chatbot that puts Google's models next to the latest from OpenAI, Anthropic, xAI, and DeepSeek.