The best current model goes from “experimental” to “previous”


As if they were the Road Runner and the Coyote, Google these days ago was able to expand its new function of editing images to match the experience that was Chatgpt’s trend with its images with Ghibli style. Now Deep Chatgpt Research It has been improved to solve consultations related to programming, and in the same afternoon Google spent the best current model for creative writing and programming of “experimental” to “prior”.

Deep Research of Chatgpt is responsible for conducting deeper research on the Internet and is another experience that Google also offers Gemini. What Openai has done now is to improve Chatgpt’s ability to integrate the First connector for Deep Research.

This connector, according to Techcrunchcreate a DIRECT LINK WITH GITHUB (in Beta) that allows developers to consult the engineering documents and code base. This connector, for the moment, will only be available for the subscriptions of Chatgpt Plus, Pro and Team in the next few days, while Enterprise and Edu is expected to arrive soon as an Openai spokesman will maintain.

And right on the day of the ad, Google has passed the Models 2.5 Flash and 2.5 Pro From “experimental” to “previous” in mobile and desktop apps of the free version and users with the Advanced subscription.

There are two important differences between them. Gemini 2.5 Flash is mainly dedicated to faster tasks that can be executed in high -connection mobile phones, while Gemini 2.5 Pro It is the model of AI that, with the new update of days ago, has become the favorite of many programmers for their ability to “chop code”.

The chatgpt github connector

Techcrunch

Now, Gemini, clicking on the model selector, show all this list with 2.5 flash and 2.5 pro updated to the previous and gives up to six options in total according to the consultation to be made. Then you can read the six and the objective of each one:

  • 2.0 flash: fast and complete help.
  • 2.5 Flash (previous): The reasoning model designed for speed.
  • 2.5 PRO (Previous): Reasoning, Mathematics and Code.
  • Deep Research with 2.5 Pro: Deep reports on the Internet.
  • 2.0 flash with search history: custom user.
  • I see 2: Generate videos from text.

When the two models enter the previous one, the day of the final version is approaching. At the moment it has only been a change in the denomination without affecting the power or capacity of both, so it is time to wait for the next announcement of the final version of Gemini 2.5 Pro.

A week in which Openai and Google have been in charge, one of Improve programming capabilities of ChatGPT, and another to step forward in a model such as Gemini 2.5 Pro that has surprised its own and strangers when reaching these days ago, with the new special edition I/O, the first position in programming in programming in Webdev Arena Benchmarks.

Leave a Reply

Your email address will not be published. Required fields are marked *