Choose image, video, audio, and research AI Tools

The catalog includes AI Tools and workflows for many marketing jobs: visuals, motion, voice, audio, research, web extraction, text generation, ads, reviews, and social data. Category filters help you narrow that library by the kind of work you want to do.

Use this article when you know the output or source material you need, but you are not sure which catalog area to start with.

Start with the outcome

Before choosing a tool, write down the job in plain language:

• "I need an image for a product launch."
• "I want to animate an existing image."
• "I need a voiceover or transcript."
• "I want to research competitors, market context, or public web data."

Then open Catalog and use the category filters to narrow the grid.

You can combine category filters with search and sort. For example, select a visual category first, then search for a model, format, or task phrase.

Choose visual categories for image work

Start with image categories when your main output is a visual asset.

Use Text to Image when you want a new image from a prompt. Use Image to Image when you are starting from an existing image and want a transformed version. Use Image Manipulation for more specific edits such as background, lighting, product placement, or object changes. Use Image Upscaling when your goal is higher-resolution output.

Open the AI Tool detail page before running it. Check the input fields, example use cases, and expected output so you know whether the tool needs a prompt, an image URL, a reference image, or another source.

Choose video categories for motion

Use video categories when you need animation, motion, avatar video, or a video cleanup step.

Use Image to Video when you have a still image and want to animate it. Use Text to Video when the prompt is the starting point. Use Avatar Video for talking-avatar or presenter-style workflows. Use Audio to Video when audio drives the motion or lip sync. Use Video Processing when your existing video needs enhancement or upscaling.

Video workflows can take longer than simple text or image tasks. Review the tool's input form and output expectations before you run it, especially when the workflow needs a source image, audio clip, prompt, or brand direction.

Choose audio and voice categories for sound

Use audio categories when the main result is speech, music, transcription, timing data, or voice direction.

Use Text to Speech when you want spoken audio from text. Use Speech to Text when you need a transcript or speech analysis from audio. Use Text to Audio for generated audio such as music or dialogue-style output. Use Voice Design when you want to create or shape a reusable voice direction.

If you plan to reuse a voice, review the output carefully before using it in a campaign or customer-facing asset.

Choose research and web categories for source material

Use Research when you need search, synthesis, market context, competitor context, company information, or AI-visibility analysis. Use Web Extraction when you already know a public page, profile, store, review page, ad library, or website source that you want the tool to collect or structure.

Public-data tools depend on what the source site makes available at the time of the run. Treat results as starting material for research and planning, then review important facts before using them in external work.

When to use search instead

Category filters are best when you know the kind of work. Search is better when you already know the tool name, provider, data source, or exact task.

Good search terms include:

• A format: "carousel", "avatar", "voice", "transcript", "image".
• A source: "LinkedIn", "TikTok", "Google Ads", "Amazon", "reviews".
• A job: "competitor", "profile", "upscale", "scrape", "research".

If the first result is not right, open a few AI Tool detail pages and compare the input requirements, output descriptions, and examples.

Where to go next

To learn the full catalog layout, read Browse the AI Tools catalog. To run a selected tool, read Run an AI Tool. For source-site caveats, read How public-data tools work and their limits.