Google has announced its latest update to the Gemini AI model, focusing all of its energy on one particular area. To keep up with the competition from OpenAI, Gemini is now an absolute coding machine.
This latest update is called Gemini 2.5 Pro Preview (I/O Edition). It’s a mouthful, but surprisingly makes sense. I/O is Google’s yearly conference where they announce the latest projects they are working on, and it's happening in just two weeks time.
While Google will almost definitely reveal further Gemini updates at this event, this is a glimpse into the direction it is heading.
According to Google, this model update preview is focused on coding and building interactive web apps. It is better at coding-based prompts across the board, but more specifically, code transformation—the ability to modify a piece of code to fit a specific goal—and editing existing code.
Announcing the launch of this latest update in a blog post and on X, Google stated that this model tops the WebDev Arena Leaderboard. This is a benchmark that focuses on an AI model’s ability to create aesthetically pleasing and functional apps.
Very excited to share the best coding model we’ve ever built! Today we’re launching Gemini 2.5 Pro Preview 'I/O edition' with massively improved coding capabilities. Ranks no.1 on LMArena in Coding and no.1 on the WebDev Arena Leaderboard.It’s especially good at building… pic.twitter.com/9vRaP6RTToMay 6, 2025
While this update won’t be of much interest to the average person developers can start building using the tool in the updated Gemini 2.5 Pro in the Gemini API via Google Studio.
It can also be used in the Gemini app, where it powers features like Canvas. Through this, anyone can vibe code and build interactive apps.
Vibe coding is a growing trend in the AI world where users can create apps and websites without needing to know how to code, focusing on the use of prompts instead.
What does this update mean?
In the world of AI, there are a few tasks that the big-name companies are trying to race to the top for. These include AI image and video generation, the ability to think through complicated tasks, and coding.
Coding through AI has seen a huge uptick in ability. xAI and OpenAI are both pouring money into it, and so is Gemini.
Recently, OpenAI spent a whopping $3 billion to buy Windsurf, an AI coding company. xAI’s Grok also saw a major boost to its coding abilities in its latest update, focusing on the ability to help users with vibe coding.
As AI becomes more accustomed to certain tasks, coding seems to be the next big barrier, and that barrier is already quickly falling.
More from Tom's Guide
- Keep your ChatGPT data private by opting out of training — here's how
- I asked Kling AI to generate a music video of my favourite band — here's how it went
- Google's Gemini AI will soon be accessible to kids under 13 — here's how that could look
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
