37 AI patterns are based on our 200+ real-world examples,
organized across five phases of an interaction journey.
Phase 1
How do users discover what AI can do in the first interaction?

Jumpstart user engagement with smart and context-aware preset actions.
Invisible AI moments that surprise users with exactly what and when they need.
Moving from clicks to natural language searches and applying filters.
Anticipating & completing actions without disrupting user's workflow.
A disclosure shown to users that clarifies AI's involvement, limitations, and sets expectations.
"For my daily work, conference talks and online courses, it’s invaluable."
Phase 2
How does the user input context into the AI?

Enabling users to interact with their voice, have a conversation or take actions.
Allowing users to write and receive information in natural handwriting.
Let users attach images, screenshots, or other visuals as part of the input.
Using physical gestures (swipes, pinches, circles) as expressive input signals for AI systems.
Offer lightweight structure (slots, hints) to compose clearer prompts faster.
Guiding users to express their intent effectively via suggestions, follow-ups, templates etc.
Let users pick the right model or mode for the task when it matters.
Let AI directly interact with other apps and services to perform tasks.
Use organizational knowledge (databases, docs, wikis) to specialize responses.
"I've gotten a ton of value out of aiverse over the last year. Patterns book was solid!"
Phase 3
How does the AI respond and in what formats?

The "test-drive" of AI to get a glimpse of results before committing to the process.
Generate or assemble video for dynamic explanations and demos.
Return images or generated visuals as the primary output form.
Speak results aloud for accessibility and hands-free use.
Presenting multiple alternative outputs for the same input.
Combine text, images, audio, and video where appropriate for richer responses.
Condense long content into concise takeaways or executive summaries.
Consistently formatting AI responses into predefined schemas instead of free-form text.
Run mutliple processes / agents concurrently to become architects of processes.
Gradual reveal of AI output, showing the progress instead of waiting blindly.
Show progress through distinct steps so users understand what's happening.
Showing how sure the AI is to help users gauge reliability, and understand the basis of the output.
Showing how the AI constructed the answer (sources, references) to build user's trust.
"Absolutely love the product and I use it at work to share with coworkers!"
Phase 4
How does the user edit, review, or improve results?

Enabling edits on specific parts of textual content, leveraging existing mental models.
Enable users to refine visually and continuously, without back-and-forth prompts.
Continue the conversation to refine results with follow-up instructions.
Request a fresh attempt with the same or tweaked prompt.
Allowing users to step through AI-suggested changes and refine them before finalizing.
Phase 5
How does the system adapt, remember, and improve over time?

AI system remembers and recalls user context, or history to support continuity and personalization.
Capturing quick sentiment on results to steer quality.
Tailor tone, suggestions, and defaults based on user history and choices.
"I genuinely found it really helpful while designing in a recent project!"

From ChatGPT to Figma AI, explore the best AI UX patterns from leading products.