Gemini 3.5 Flash now supports the use of computer tools. This functionality allows the model to interact with external applications and services. The model can execute code, access databases, and utilize other tools to achieve specific goals. This expands the scope of tasks the model can handle.
This capability is enabled through a system of tool calling. The model identifies when a tool is needed and then invokes the appropriate tool with the necessary parameters. The tool executes its task and returns the results to the model, which then uses the results to continue its reasoning process.
This approach provides a structured way for the model to leverage external resources. The system is designed to be flexible and extensible, allowing for the addition of new tools and capabilities over time. The model’s performance is evaluated based on its ability to solve complex problems that require the use of multiple tools.
Source: https://deepmind.google/blog/introducing-computer-use-in-gemini-3-5-flash/