During 2023 and 2024, Google gave the feeling of going behind in the advances in artificial intelligence for the consumer market against competitors such as Microsoft And above all, OpenAI. This year, however, it is being able to reverse that perception thanks to the deployment of Gemini in its platforms and applications, as well as by the novelties that are incorporated, with special mention to those presented in the last Google I/O. of last May.
This Monday, the company has re -updated its offer of language models within the chatbot, but it has also included a very relevant novelty that has not announcedbut they have detected in 9to5google.
The Multimodality of the AI, which allows them to accept different data formats such as entry and exit, has taken a new step forward with Gemini’s ability to process videos provided by the user, analyze them and respond on them.
Why has Google not announced an improvement of this draft? From the middle they point out that the function It is not yet available in a generalized way and varies according to accounts and devices. They can use both free and payment accounts in Android and iOSbut it is not yet available on the web version. Presumably, for now, only in the US market. It is more as if it were in evidence than being a formal launch, which would explain that Google has not said anything.
In the tests performed, 9to5google obtained descriptions on the videos uploaded to the chatbot, which also He showed them inside a player along with the answer.
This is an important improvement on multimodality that they currently offer so much Chatgpt as COPILOTthat probably will not take long to incorporate that capacity. Logically, that a user can upload any video to the platform, which predetermined them uses them for training as the rest of the content that is introduced, opens a new front within the issue of intellectual property and generative artificial intelligences.
The other novelties that Gemini He received on Tuesday they were a new model that is added to the family Gemini 2.5 and changes of the state of others. Gemini 2.5 Pro and Gemini 2.5 Flashannounced in the last Google I/O, pass from the Preview to Stable and are available for all users.
The new 2.5 Flash-litein the Preview and Available through the platforms Google Ai Studio and VERTEX AIis presented as the fastest and profitable family Gemini 2.5 and is designed for high volume tasks where cost efficiency is key. It is expected that not too late to reach the app of Gemini.