Home News Google claims Gemini 2.5 Pro Preview beats DeepSeek R1 Grok 3 Beta...

Google claims Gemini 2.5 Pro Preview beats DeepSeek R1 Grok 3 Beta and in coding performance

0
Google claims Gemini 2.5 Pro Preview beats DeepSeek R1 Grok 3 Beta and in coding performance

Image credit: VentureBeat via Midjourney

Join the event trusted for over two decades by business leaders. VB Transform brings the people who are building enterprise AI strategies together. Learn more

Google released an updated preview version of Gemini 2.5 Pro as a preview. The model was first announced in March, upgraded in May and will be available in general in a few weeks.

Enterprises are able to test new applications or replace older versions with a version of Gemini 2.5 Pro’s “I/O Edition” that is updated. Google’s blog post is more creative and performs better in coding and reasoning than other models.

The latest Gemini 2.5 Pro version is now available in preview.

It is better at coding and reasoning, science + mathematics, and shows improved performance on key benchmarks. @lmarena_ai with a 24pt Elo jump since the previous versions.

Also… pic.twitter.com/SVjdQ2k1tJ

– sundar pichai (@sundarpichai)””https://twitter.com/sundarpichai/status/1930656033237823862?ref_src=twsrc%5Etfw””> Google will announce on June 5, 2025 that they have updated Gemini 2.5 Pro, making it better than the previous iteration. This was quietly released in May. Demis Hassabis, CEO of Google DeepMind, said that the I/O version is the best coding model the company has ever released.

This new preview, named Gemini 2.5 Pro Preview 06-05 Thinking is even better than I/O Edition. The stable version Google plans to release publicly is “ready for enterprise-scale capabilities.”

The I/O edition, or gemini-2.5-pro-preview-05-06, was first made available to developers and enterprises in May through Google AI Studio and Vertex AI. Gemini 2.5 Pro 06-05 Thinking is also available via the same platforms.

Performance metrics

The new version of Gemini 2.5 Pro is even more efficient than the previous release.

Google stated that the new version Gemini 2.5 Pro has improved by 24 points on LMArena. It also improved by 35 points on WebDevArena where it is currently ranked first. The benchmark tests conducted by the company showed that this model outscored its competitors. Openai ‘s o3, O3-mini and o4 Mini. Anthropic’s Grok 3 Beta, Claude 4 Opus from Accounts DeepSeek R1.

Google stated in a blog post that “We’ve also addressed the feedback from our previous 2.5 Pro release, improving its structure and style — it can be creative with better-formatted answers.”

What enterprises can expect.

Google’s continuous improvement of Gemini 2.5 Pro may be confusing to many, but Google has previously framed this as a response community feedback. The new version costs $1.25 for each million tokens, without caching, and $10 for outputs.

Matt Marshall, VentureBeat, called the first version of Gemini 2.5 Pro “the smartest model that you’re not currently using” when it launched in March. Since then, Google’s integrated the model into a number of new applications and services including “Deep Think,” which allows Gemini to consider multiple hypotheses prior to responding.

With the release of Gemini 2.5 Pro and its two upgraded version, Google regained its place in the large-language model space, after competitors such as DeepSeek and OpenAI diverted industry attention to their reasoning modeling.

Within a few hours after the announcement of the updated Gemini 2.5 Pro version, developers began playing with it. Although many users found that the update lived up to Google’s promise to be faster, it is still unclear if the latest Gemini 2.5 Pro actually performs better.

The first hour with “Gemini 2.5 Pro Preview 06-05”

Before: “You are absolutely…

— Patrick Bade (@nishffx) June 5, 2025

You guys cooked and really enjoyed the app builder.

created a game, tested it and used imagen to create assets on the fly? It’s hosted, it’s easy to share. The best no-code builder I’ve seen.

Keep building out the vibe App Marketplace, this could… June 5, 2025

Gemini 2.5 Pro Preview was pretty good. I used it yesterday to do deep research, and the results were better than some of those big names.

– Janak (@janaks09).””https://twitter.com/janaks09/status/1930697398252138574?ref_src=twsrc%5Etfw””> June 5, 2025

VB Daily provides daily insights on business use-cases

Want to impress your boss? VB Daily can help. We provide you with the inside scoop about what companies are doing to maximize ROI, from regulatory changes to practical deployments.

Read our privacy policy

Thank you for subscribing. Click here to view more VB Newsletters.

An error occured.

www.aiobserver.co

NO COMMENTS

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Exit mobile version