AI models for what you want to make
Explore video, image and chat models. Choose the inputs, output format and controls your task needs.
Image
Video
Research
- Claude Opus 5Explore a topic in depth and build detailed presentations.
- Claude Opus 5.5Refine complex ideas into clear writing that follows your style.
- Claude Sonnet 5Research a topic and turn your findings into a plan.
- Claude Sonnet 5.5Polish drafts and presentations with quick, natural revisions.
- Gemini 3.1 ProMake complex ideas clear with visual explanations and creative prototypes.
- Gemini 3.6 FlashGet quick feedback on images, layouts, and visual ideas.
- Gemini 3.8 FlashPlan ambitious projects and work through questions step by step.
- GPT-5.6 LunaQuick answers and short text edits to keep you moving.
- GPT-5.6 SolCareful reading and analysis of long documents.
- GPT-5.6 TerraTalk through ideas and shape your next draft.
- GPT-6.1 SolCarry a detailed brief through a long writing or research task.
- GPT-6 AstraDeep research and thoughtful judgment for your most complex ideas.
About
Choose an AI model for the work in front of you. Video models offer different clip lengths, input methods, resolutions and sound controls. Image models let you generate a new picture or edit an existing one, with different shapes and output settings. Chat models handle the Agent conversation as you write, investigate or plan. Open a model page to review its options, or compare two models when the choice depends on a particular setting. Keep Auto when you want HowWorks to choose a supported model.
Questions
How do I choose between AI video, image and chat models?
Start with the result you need. Choose a video model for a moving scene, an image model for a picture or edit, and a chat model for the Agent conversation. Within each group, compare the supported inputs and controls that matter for your task.
What does Auto do?
Auto lets HowWorks choose a supported model for the request. Select a named model when you want to work with that model and its available settings. Model pages keep the named model selected.
Is the chat model the same as the model that creates an image or video?
They are separate choices. The chat model handles your conversation with the Agent; an image or video tool uses its own generation model. You can discuss the idea in chat before asking the appropriate tool to create it.
What do Standard, Fast, Mini and Turbo mean?
They are tier names offered by particular models, with different supported settings. The same tier name does not establish the same resolution, input support or speed across models. Check the selected model’s options before generating.