> ## Documentation Index > Fetch the complete documentation index at: https://docs.imagine.art/llms.txt > Use this file to discover all available pages before exploring further. # Getting Started > The interface, the mode pills, the AI Images and AI Videos presets, and the two controls that shape every task: model tier and Ask permission. ## The sidebar Go to **imagine.art/computer**, or choose **Computer** from the workspace switcher at the top of the ImagineArt sidebar. You'll land on the main chat view. * **New Chat**: start a fresh conversation. * **Search Chats**: find a past conversation. * **Artifacts**: a library of everything your chats have produced, filterable by Slides, Websites, Documents, Sheets, Images, and Audio & Video. * **Startup**: the website builder. Describe a game, 2D or 3D site, or landing page and publish it. * **Integrations**: [connect your tools](/imagine-computer/connectors). * **Knowledge**: your team's files and data, which you can use in chats. * **Projects** and **Recents**: your saved work and chat history. ## Modes: presets on the same chat The pills below the prompt box (**Create Slides, Startup, AI Images, AI Sheets, AI Docs**, and under **More**: **Video Editor, AI Videos, AI Explainer, Graphs, AI Music, Deep Research, AI Resumes, Plan Trips, Motion Design, AI Contracts, Chat, AI Ads**) mostly aren't separate tools. Each one (other than Startup) is the same chat framework on its own sub-route, with a mode-specific composer hint and starter prompts. **AI Images** and **AI Videos** go further, with their own controls and template galleries. See [Start from a preset](#start-from-a-preset-in-ai-images-and-ai-videos). Imagine Computer home with the More menu open, listing Video Editor, AI Videos, AI Explainer, Graphs, AI Music, Deep Research, AI Resumes, Plan Trips, and Motion Design **Startup is the exception.** Clicking the **Startup** pill opens the separate website builder, with its own **Build** composer, rather than a chat preset. ## Two controls that shape every task Every prompt bar has two controls worth understanding before you run anything. ### Model tier Model tier dropdown open below the prompt box, showing Flash, Pro, and Ultra with their descriptions * **Flash** — fast and efficient for everyday tasks, quick ideas, and lightweight problem-solving. * **Pro** — more power for deeper research, bigger workflows, and multi-step tasks that need extra capacity. * **Ultra** — maximum intelligence for your most complex, high-stakes, end-to-end work. ### Ask permission Plus menu in the prompt box showing Upload Photos & Files, Web search, Personalization, Ask permission, and Use commands Click **+** in the prompt box to find the **Ask permission** toggle, alongside **Upload Photos & Files**, **Web search**, **Personalization**, and **Use commands**. * **Off** (default): the agent runs without stopping, with full access to create, edit, or run intensive actions. * **On**: the agent is meant to ask before creating or editing artifacts. In earlier testing, the approval setting did not actually pause a full research run at any point, including at the report-writing step. Don't rely on it as a safety gate. Treat every run as if it will continue without stopping. ## Start from a preset in AI Images and AI Videos The **AI Images** and **AI Videos** homes let you start from a ready-made look instead of a blank prompt. Open them from the **AI Images** pill, or from **More** > **AI Videos**. Below the prompt box, four dropdowns control the output: * **Model**: pick a model, or **Auto** in AI Images to let Computer choose. * **Settings**: in AI Images, set **Quality** (1K, 2K, 4K), **Aspect Ratio**, and **Variations** (1, 2, or 4). In AI Videos, set **Quality** (480p, 720p, 1080p), **Aspect Ratio**, and **Duration**. * **Style**: apply a visual preset. AI Images includes 20 styles, such as **Photorealistic**, **Anime**, **Cinematic**, **Ghibli**, **Pixel Art**, and **Watercolor**. * **Camera**: choose a camera angle. AI Images offers **Birdseye**, **Dutch**, **Low Hero**, **Medium**, and **Wide**. AI Images home with the Images prompt box, model, settings, Style, and Camera dropdowns, Featured Templates, and Templates category filters Browse **Featured Templates**, or filter the **Templates** gallery by category. Hover a template and click **Use Template**. | Home | Template categories | | - | - | | **AI Images** | Ad Creatives, Product Shots, Style Transfer, Interior Design, Fashion Shoot, Social Thumbnail | | **AI Videos** | Trending, AI Ads, Viral Videos, Cinematic Shot, Product Demo, UGC, Music Video, Fashion Film | The template opens in its own window with example results, a prompt that's already written for you, and the inputs it needs. For example, **Product Shots** asks you to **Add Product**, lets you change the **Style Reference**, and has controls for the number of images, aspect ratio, and quality. Edit the prompt if you like, then click the create button (for example, **Create Product Shot**). Product Shots template window with example images, Add Product upload, Style Reference dropdown, editable prompt, count, aspect ratio, and quality controls, and a Create Product Shot button The **AI Videos** home works the same way, with video models, video settings, and video templates such as **AI Ads**, **Viral Videos**, **Cinematic Shot**, **Product Demo**, **UGC**, **Music Video**, and **Fashion Film**. AI Videos home with the Videos prompt box, Seedance 2.5 model, 480p / 16:9 / 4s settings, Style and Camera dropdowns, Featured Templates, and Templates category filters This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.