The revamped version of Gemini 2.5 Pro is the smartest model and the foremost programming expert from Google

Google has advanced its plans and has just launched its new version of Gemini 2.5 Pro with enhanced capabilities and particularly excels in programming. The tech giant had planned to release this update during its I/O event to be held on May 20 and 21, but ultimately decided not to make its users wait.

The updated version of Gemini 2.5 Pro is now accessible in the Gemini API via Google AI Studio and Vertex AI, and from the Gemini app.

We just dropped Gemini 2.5 Pro (I/O edition). It is our most intelligent model that is even better at coding.

Now, you can build interactive web apps in Canvas with fewer prompts.

Head to https://t.co/BCZTezi8r9 and select “Canvas” in the prompt bar to try it out, and let us know… pic.twitter.com/gjDpf5nkRE

— Google Gemini App (@GeminiApp) May 6, 2025

By selecting the new version of Gemini 2.5 Pro in the Gemini app, this model enhances the capabilities of the Canvas function and will allow programming of interactive web applications with fewer instructions.

Programming is simplified with the new Gemini 2.5 Pro

Google states that its revamped Gemini 2.5 Pro is its most intelligent model and the best in terms of programming. Thanks to this AI, the creation of interactive web applications becomes a very simple and accessible process.

The tech company claims that, due to the applied enhancements, this model now leads the WebDev Arena rankings (which measures a model’s ability to create aesthetically appealing and functional web apps), surpassing its previous version by 147 Elo points.

In addition to this, Google asserts that the new Gemini 2.5 Pro “continues to develop its strong foundation in native multimodality and long-context; it offers cutting-edge performance in video comprehension, with a score of 84.8% on the VideoMME benchmark.”

Demis Hassabis, Co-Founder & CEO of the Google Deep Mind AI division, highlighted its utility for testing ideas and creating prototypes.

Very excited to share the best coding model we have ever built! Today we are 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 is especially good at building… pic.twitter.com/9vRaP6RTTo

— Demis Hassabis (@demishassabis) May 6, 2025

Meanwhile, Dave Messer, Product Manager of Learning Labs and Gemini Learning, demonstrated how he created a sound-guessing game.

🔊 sound on! Check out this fun sound-guessing game I made using Gemini 2.5 Pro! It is a game where you try to guess the sound by drawing! It was super easy to create with Canvas.
Try it out here: https://t.co/NJCykdEbQr@GeminiApp @GoogleDeepMind @Google @googledevs pic.twitter.com/l48dgIroDQ

— Dave Messer (@davemesserx) May 6, 2025

In the same vein, numerous users have rushed to try this novelty and have shared their excitement about the results achieved.

Wow Gemini 2.5 Pro has received a big update

This new model generated a full 3D city simulation in a single prompt 🔥

Its code and UI generation capabilities are impressive.

You can already use it for free pic.twitter.com/2SGy4HdVI6

— Paul Couvert (@itsPaulAi) May 6, 2025

I have tried the new version of Gemini 2.5 in programming to see if it was as good at coding as its advertisement says…

And as they announced, the design in the user interface I believe is a good step forward

Tool to visualize rankings, it took me a few hours.. 🧵👇 pic.twitter.com/ro0Bi4ZaBi

— Lino Uruñuela (@Errioxa) May 7, 2025

Photo: Google