Logo La Derechadiario US
ES
Logo X
Gemini 2.5 logo next to a futuristic illustration of an android with technological details and the Google logo on its ear
Flash-Lite: Google's new AI that competes in price and speed | Camara La Derecha Diario
ARGENTINA

Google launches Gemini 2.5 Flash-Lite: its fastest and most affordable AI for large-scale tasks

Google introduced Flash-Lite, its new AI model that combines speed, efficiency, and low cost for large-scale tasks

by

Editorial Team

Google  introduced the final version of its artificial intelligence Gemini 2.5 Pro and Flash models, and also announced the preview of Gemini 2.5 Flash-Lite.

This new version stands out for its speed, low cost, and excellent performance in translation, classification, and reasoning tasks.

Laptop screen displaying the homepage of Gemini, Google's artificial intelligence, with a blue button to start the chat and an image of a person smiling next to explanatory text.
Google unveiled the final version of its Gemini 2.5 Pro and Flash artificial intelligence models | La Derecha Diario

What is Gemini 2.5 Flash-Lite and why does it matter?

Flash-Lite is the new fastest and most efficient AI model in the 2.5 series. It outperforms previous versions and is designed for high-volume tasks with low latency.

It positions itself as an ideal option for developers seeking performance and low cost. It is also the fastest "2.5" model Google has launched so far.

What sets Gemini 2.5 Flash-Lite apart from other models?

Compared to 2.0 Flash-Lite and 2.0 Flash, this model improves in every aspect:

  • Faster and more accurate coding
  • Deeper mathematics and reasoning
  • Multimodal input with 1 million context tokens
  • Higher response speed (tokens per second)
Bar chart comparing the token generation speed per second of different artificial intelligence models from Google, OpenAI, Anthropic, DeepSeek, and xAI
What sets Gemini 2.5 Flash-Lite apart from other models? | La Derecha Diario

It also includes key tools such as connection with Google Search, code execution, and activation of deep reasoning, even on limited budgets.

How does Flash-Lite compare in costs to other models?

Usage cost is one of its major advantages. While Gemini 2.5 Pro costs 1.25 dollars per million tokens, Flash-Lite is just 0.10 dollars.

Comparative table of prices and performance of different versions of Gemini 2.5 in reasoning, science, mathematics, code generation, and code editing tasks
The cost of use is one of its main advantages | La Derecha Diario

This is a key difference for those offering services to third parties and needing to scale without breaking their budget.

What tasks does Flash-Lite do best?

This new model excels at tasks such as:

  • High-volume translation
  • Real-time content classification
  • Video compression and multimodal processing

Google showed charts where Flash-Lite produces many more tokens per second than GPT-4.1 and other models on the market.

Comparative table of results from different artificial intelligence models across various benchmarks in visual, auditory, and subtitle modalities, showing the scores achieved by Gemini 1.5 Flash, Gemini 1.5 Pro, Gemini 2.0 Flash-Lite, Gemini 2.0 Flash, Gemini 2.5 Flash, Gemini 2.5 Pro, and OpenAI GPT 4.1.
What tasks does Flash-Lite perform better? | La Derecha Diario

Where is Gemini 2.5 Flash-Lite available?

The model can already be tested in Google AI Studio and Vertex AI. Customized versions are also being added for tools such as Google Search and generative AI mode.

The final versions of Gemini 2.5 Pro and Flash are already available from the Gemini app.

➡️ Argentina

More posts: