The Yandex team has released an update to the desktop version of its browser, introducing an ML model (machine learning model) to optimize RAM consumption. Artificial intelligence identifies inactive tabs that the user will not need in the near future and temporarily unloads them from memory.
According to the company's measurements, after the introduction of the model, "Yandex Browser" consumes up to 36% less RAM than Google Chrome. This means that simultaneously running programs — office applications, games, or messengers — will work faster and more stably. Currently, the browser frees up to 1 GB more memory than before the ML model appeared.
The model takes into account dozens of parameters: the state of the tab (whether video or music is playing), the number of open tabs, the frequency of switching between them, and even the distance between different tabs. When a unloaded page is reopened, it is reloaded, ensuring continuity of work.
The user can manage the function in the "Browser" settings, in the "Performance" section. The ML model is active when the "Always unload unused tabs from memory" option is enabled.
