AI Studio

Open AI Studio, type your question in the message box, and press Enter. AI Studio creates a thread and streams the reply.

Select the model name at the top of a thread, pick a group, then pick a model. A model picked from a provider group becomes your default.

The model catalogue is served at run time and differs by Account. This page carries the model facts that live in the code.

Attach files from your device or from Drive with the plus button next to the message box, or drag them onto the thread.

Attach up to five GitHub repositories to a thread so the model reads their file listing and README as reference on every message.

Select the persona control at the top of a thread and choose a persona. Its text is added to the model's instructions on every message you send.

Open Personas in the AI Studio sidebar, select Add Persona, and fill in the role and instructions the model should follow.

Use the plug button next to the message box to add a remote MCP server so the model can reach its tools and data during the thread.

Turn on Image in a thread to get pictures instead of text, or open Media in the sidebar to generate images and videos with size and quality controls.

The Voice button next to the message box holds two modes: Live Chat, a spoken back-and-forth, and Dictate, which types what you say into the box.

Choose Transcribe audio from the plus button next to the message box, pick a recording, review the transcript, then insert it into the thread.

Recent threads are listed by date in the left sidebar. For older ones, open All threads and search, filter or sort the full list.

From eight messages on, a conversation map appears at the right edge. Open it to jump to any message, and collapse your own prompts to see replies only.

Select Fork on any message to start a new thread that holds the conversation up to that point, leaving the original untouched.

Use the More options menu at the top of a thread to rename it or delete it, or delete several at once from the Threads page.

Export a thread as a Word document or a text file, copy it to the clipboard, or copy a link to it, from the More options menu.

Open Agents in the AI Studio sidebar, pick an agent, describe the work, fill in the fields it asks for, and select Start a run.

Agent Executions on the Agents page lists every run. Open one to see its steps and outputs, stop it while it is going, or start it again.

Every thread has a token budget set by its model. The counter beside the message box shows how much is used, and the thread stops at the limit.

The Plans page in the AI Studio sidebar shows what pays for your requests, which limits can stop them, and a history of what you have used.

A GitHub repository, issue or pull request link in a message renders as a card with its state and counts. Other GitHub links stay plain text.

A reply that ends early carries a note saying who ended it. The note tells you whether to send again, lower the reasoning level, or change model.

When AI Studio offers no model, or refuses the one you chose, the cause is your Account's model policy, your plan, or the model's own limits.

AI Studio refuses a file when it is too large, when the model cannot read that type, or when the upload itself failed. Each case names itself.