macOS generative visuals, local-first

Describe the visual. Prismo handles the canvas.

Prismo is a local-first macOS app for generative visuals, audio-reactive scenes, webcam inputs, and chat-driven canvas control. Support, privacy details, and App Store-ready pages live here so operators know exactly how the app behaves before they turn on a permission or an AI mode.

Open Support Read Privacy Policy

macOS 14+ runtime

Audio + webcam inputs

Optional AI modes

No tracking by default

Inside The App

A control room for live visuals

Prismo separates the control surface, the full canvas, and runtime settings so you can steer the visuals without losing the show.

Starter looksLive

Dark Club Visual

Ready

Calm Ambient Scene

Ready

Set Up Audio Sync

Ready

Find a Starter Look

Ready

Inputs

Audio, webcam, photos

Agent mode

Optional, never required

Why Prismo

Built for creative control without a cloud-first trap

The app keeps rendering local while still giving you room to react to audio, camera, photo, and agent-assisted workflows when you want them.

Local-first creative runtime

Prismo keeps rendering on your Mac so your visual workflow stays responsive, private, and close to your files.

Reactive inputs on demand

Build visuals from microphone audio, app or system audio capture, webcam input, and photos you explicitly choose.

Chat-guided scene control

Describe what you want to change and let the Prismo agent help shape scenes, parameters, and looks.

Multiple visual engines

Move between WGSL, GLSL, p5, Three.js, Babylon, and more without flattening the creative workflow into one renderer.

How It Works

Start with intent, then tune what the scene reacts to

Prismo is designed around one simple pattern: pick a look, choose the input, and move the result to the dedicated canvas window.

Start from a look

Open a preset like Dark Club Visual or Calm Ambient Scene, or ask the agent to build a new direction from plain language.

Dial in the input

Choose whether the scene reacts to microphone audio, app or system audio, webcam input, or photos you select.

Push it to canvas

Run visuals in the dedicated canvas window, keep control in the main window, and use settings only when you need runtime or permission changes.

Permissions

Only the permissions the scene actually needs

Prismo requests access feature-by-feature. If you never turn on webcam, microphone, photo, or app-audio flows, those permissions stay unused.

Camera

Used only when you enable webcam-reactive visuals.

Microphone

Used only when you enable microphone-driven audio reactivity.

Photos

Used only when you choose photo-based visuals or ask Prismo to save captures.

Screen + System Audio Recording

macOS may request this when you enable app or system audio capture so Prismo can browse or capture that audio.

Support At A Glance

Need help before or after App Store review?

The support page covers troubleshooting, required bug-report details, and the quickest way to reach a human.

Support email

kelly@lobsterwork.ai

Share this in bug reports

Your Prismo version and macOS version.
What scene or input you were using.
The exact steps that reproduced the issue.
Any screenshots, screen recordings, or crash text you can include.

Open Support Read Privacy Policy

FAQ

Straight answers for support, privacy, and setup

Does Prismo upload my photos or media by default?

No. Prismo is built around a local-first workflow. It accesses your photo library only when you explicitly choose photo-based visuals, and it may save rendered captures only when you ask it to.

Do I need AI features to use Prismo?

No. The core visual runtime is local. Optional AI features are available if you enable API key mode or ChatGPT OAuth mode for the embedded agent.

What permissions might macOS ask for?

Depending on the features you turn on, macOS may ask for camera, microphone, photos, or Screen and System Audio Recording access.

Where should I go for help or App Store review questions?

Use the support page for troubleshooting and contact details, or email kelly@lobsterwork.ai with your app version, macOS version, and the steps that reproduced the issue.