We’ve seen developers doing amazing things with Gemini 2.5 Pro, so we decided to release an updated version a few weeks early to get into developers hands sooner.
Today we’re excited to release Gemini 2.5 Pro Preview (I/O edition). This update features even stronger coding capabilities, for you to start out constructing with before Google I/O later this month. Expect meaningful improvements for front-end and UI development, alongside improvements in fundamental coding tasks equivalent to transforming and editing code, and creating sophisticated agentic workflows.
“We found Gemini 2.5 Pro to be the very best frontier model in relation to “capability over latency” ratio. I sit up for rolling it out on Replit Agent at any time when a latency-sensitive task must be completed with a high degree of reliability.”
– Michele Catasta, President, Replit
Best-in-class frontend web development
Gemini 2.5 Pro now ranks #1 on the WebDev Arena leaderboard, which measures human preference for a model’s ability to construct aesthetically pleasing and functional web apps. Drawing on this leading capability, Gemini 2.5 Pro powers Cursor’s modern code agent and empowers our collaborations with firms like Cognition and Replit. Together, we’re pushing the frontiers of agentic programming to unlock recent possibilities for developers.
“The updated Gemini 2.5 Pro achieves leading performance on our junior-dev evals. It was the first-ever model that solved certainly one of our evals involving a bigger refactor of a request routing backend. It felt like a more senior developer since it was capable of make correct judgement calls and select good abstractions.”
– Silas Alberti, Founding Team, Cognition
Gemini 2.5 Pro in motion
Gemini 2.5 Pro’s deep understanding of code, combined with powerful reasoning, continues to make Gemini 2.5 Pro the go-to model for developers. We’re particularly enthusiastic about how this model might be utilized in the next cases.
Video to code
Gemini 2.5 Pro delivers state-of-the-art video understanding, scoring 84.8% on the VideoMME benchmark. Combining this with coding enables recent flows that were previously impossible with previous versions. For instance, the Video to Learning App in Google AI Studio demonstrates how Gemini 2.5 Pro creates an interactive learning app based on a single YouTube video. With improved video understanding and complete UI, the updated Gemini 2.5 Pro model delivers a more functional experience than the previous easy example.
Easier feature development
Gemini 2.5 Pro is powerful at front-end web development, helping you get more done. Implementing recent features means manually diving into design files and inspecting components to match style properties like colours, fonts, padding, margins, and borders then manually writing the CSS code needed to copy those visual properties accurately. Now imagine using Gemini 2.5 Pro in an IDE and having the model generate recent features, like adding a video player within the sort of the opposite apps within the Gemini 95 starter app.
Quick concepts to working apps
Bringing ideas to life with each functionality and a good looking UI is made easier with Gemini 2.5 Pro. The brand new dictation starter app, built using the updated model, is an amazing example of this in motion. Concentrate to a few of the details just like the wavelength animations, responsive design, and subtle button hover effects. By default, the model has an actual taste for aesthetic web development while maintaining its steerability, helping developers quickly take an idea to a working web app. Gemini 2.5 Pro was capable of design and code the microphone UI animation for the dictation starter app.
Start constructing with Gemini 2.5 Pro
You possibly can construct with Gemini 2.5 Pro with the Gemini API in Google AI Studio, and enterprise customers can use Vertex AI.
For developers already using Gemini 2.5 Pro, this new edition is not going to only improve coding performance but will even address key developer feedback including reducing errors in function calling and improving function calling trigger rates. The previous iteration (03-25) now points to essentially the most recent version (05-06), so no motion is required to make use of the improved model, and it continues to be available at the identical price. We have now also updated the model card with the new edition of two.5 Pro .
We will’t wait to see the amazing apps you construct!
