The Path of AI Creator News: Google announces Gemini 2.5 Flash-Lite! With improved thinking power, the AI model has evolved even further. Check out the details! #Gemini2.5 #GoogleAI #AIModel
Video explanation
Google's new AI "Gemini 2.5 Flash-Lite" is here! Is it really smart, fast, and affordable?
Hello everyone! When you hear the word AI, some of you may think, "It sounds difficult...". However, AI has recently become more and more familiar to us. Today, we will bring you some new news from the world of AI, announced by Google! It's called"Gemini 2.5 Flash-Lite"It's a cool name that sounds strong, but what kind of AI is it? Let's take a look!
What is "Gemini" anyway? Let's review the basics of AI!
I'm sure there are some of you who have never heard of Gemini before, so I'll give you a brief explanation. Gemini is the name of a series of very intelligent AI models developed by Google."A very clever computer brain"It can learn a lot of things, answer our questions, write sentences, draw pictures, and do a lot of other things.
Google is adding new members to the Gemini family and powering up existing models. And now, a preview version (like a trial version) of "Gemini 2.5 Flash-Lite" has been announced, which will bring a breath of fresh air to the family!
New members and updates to the Gemini 2.5 family!
In fact, there are several versions of Gemini, each with different strengths and features. In this news, in addition to the new "Flash-Lite", it was also announced that the existing "Gemini 2.5 Pro" and "Gemini 2.5 Flash" have become the official versions.
Feature 1: "Thinking" AI, Gemini 2.5 Series
The main feature of this new Gemini 2.5 series is that it is AI-enabled."think"That's it! Technically speaking, it's called a "thinking model," and by taking the time to "think" carefully before giving an answer, rather than just processing information, you can give a more accurate and high-quality answer. It's amazing, isn't it?
Moreover, developers are looking to add a function that allows them to adjust the length of this "thinking time," that is, the amount of "thinking" the AI needs to do.Thought BudgetThis allows you to control the speed of AI's response and the depth of its answers to suit your needs.
New release! "Gemini 2.5 Flash-Lite" is the star of cost performance and speed!
Now, the star of the show is the Gemini 2.5 Flash-Lite! As the name suggests, this model has both Flash and Lite."Speed" and "Lightweight (Affordable)"It was created with the aim of achieving the following: It is the cheapest in the Gemini 2.5 family, and is characterized by its fast response speed (also called latency, the time it takes for the AI to respond).
You may be thinking, "But isn't this an AI that 'thinks'? Doesn't it take time to think?" In fact, Flash-Lite's "thinking" function is turned off by default. This is to make it as fast and cheap to use as possible. However, it is cleverly designed so that you can turn on the "thinking" function when necessary!
When would it be useful? For example...
- Automatically sort a large number of emails by content (Sort）
- Summarize the minutes of a long meeting (要約）
It's perfect for when you want to process a large amount of information quickly. Moreover, compared to the previous "Gemini 1.5 Flash" and "2.0 Flash", the performance has been improved and it can return the first word (called a token. It is the basic unit when AI understands words and creates sentences, and it can be a letter or part of a word) with less waiting time. This is very promising!
Developers can fine-tune this "amount of thinking" through an API (an application programming interface (API), which is like a connection port that allows apps and services to exchange information with each other).
"Gemini 2.5 Pro" and "Gemini 2.5 Flash" are now official versions!
In addition, the preview versions of "Gemini 2.5 Pro" and "Gemini 2.5 Flash" have already been released, and they are now finally available as official versions that are stable and can be used!
This is a bit technical, but it seems that the pricing structure for "Gemini 2.5 Flash" has also changed. For example, if you let the AI read information (Input Token) will cost a little more, and instead you can have AI generate the answer (Output TokenThe price for the AI has dropped a little. Also, the price used to be different depending on whether the AI was "thinking" or "not thinking," but that distinction has been abolished. It feels simpler now!
Which Gemini model is right for you?
According to Google, the recommended usage for each model is as follows:
- Gemini 2.5 Flash-Lite: For when you want to complete a large amount of work as cheaply and efficiently as possible!
- Gemini 2.5 FlashFor when you want fast performance and smooth execution of everyday tasks.
- Gemini 2.5 Pro: For when you want to write programming code, perform highly complex analysis, or outsource advanced tasks.
By the way, the Gemini 2.5 series itself is a relatively new model, first announced on March 3th of this year.
A word from the author, John
Wow, AI is evolving so quickly! With the introduction of the new Gemini family, I'm excited to think that this amazing AI technology will become easier to use by many people in many different situations. I'm looking forward to seeing what interesting things developers will create with these new AIs! I look forward to a future where AI makes our lives more convenient and enriching.
This article is based on the following original articles and is summarized from the author's perspective:
Google previews Gemini 2.5 Flash-Lite