AI Studio for Live Video — Generate Overlays, Graphics, Video & Sound

Your show's graphics department is now a prompt.

VisionMixer's AI studio turns a typed idea into a fireable graphic: stream overlays with real live controls, images and video clips that arrive already keyed, voices, sound effects and music — all generated inside the same Mac app that runs your show.

For streams · webinars · classes · worship · meetings

ClaudeGeminiVeoElevenLabsyour keys · their rates · no markup

AI graphics for live streaming, in three steps

Describe it

Type what you want in plain words: “a gold lower third with my name,” “a ten-minute countdown that turns red in the final minute,” “rain on green.”

Refine it

One click rewrites your rough idea into an engineered prompt — layout, motion, details you didn't think to spell out — before anything is generated.

Fire it

The result lands in your library and on the stage, ready to cue like anything else in your show — from the deck, a hotkey or your phone.

Overlays that are alive

AI stream overlays with real controls

Most AI tools hand you a flat picture. VisionMixer generates HTML overlays — living graphics with looping animation, editable settings and declared live controls. Ask for a countdown timer and it arrives with actual Start, Stop and Reset buttons: in the app, on deck tiles, on hotkeys, even on your phone remote.

Quick starters for lower thirds, tickers, countdowns and badges
Editable settings — change the text, colors or timing without regenerating
Live controls show up wherever you cue things: deck, hotkeys, phone

The VisionMixer Show Deck: a grid of scene tiles, sound pads and channel faders.

Green screen without the screen

AI green screen video, already keyed

Describe an image or a short video clip — 4 to 8 seconds, with audio — and VisionMixer generates it on chroma green, then delivers it to your stage already keyed. Confetti bursts, a floating product, weather over your shoulder: AI generated video for streams that layers straight over live cameras, no masking session required.

Generated images and 4–8 second video clips, with sound
Arrive on the stage keyed and ready to layer over live video
Your real camera too: AI person segmentation, no fabric involved

The VisionMixer layer inspector: scale, position, glow, keying, corner-pin, masking and effects for the selected layer.

Sound on demand

AI sound effects, music and voices for live production

Type “a deep bass swell with a metallic hit,” or hand a script to a generated voice. VisionMixer creates voices with text-to-speech, sound effects and music from plain descriptions, and files every result in the library's Sounds section and on your soundboard pads — ready to fire mid-show.

Text-to-speech voices for intros, announcements and characters
Sound effects and music described in a sentence
Results land on soundboard pads, cued like everything else

Create → Audio in the Media Studio: generating a sound effect named “Deep Base Hit” from the one-line description “a deep bass swell with a metallic hit”.

A detailed lo-fi hip-hop music prompt — engineered by the Refine Prompt button — ready to generate a 30-second stream intro track.

Deck sound pads: the generated lo-fi intro track playing, beside Deep Base Hit, Applause, Cheer, Error and Drum Roll tiles.

Your keys, their rates, no markup

Bring your own keys

Claude (Anthropic) or Gemini write overlays and text. Gemini makes images, Veo generates video, and ElevenLabs and Gemini handle voices, sound effects and music. Set your keys up once in Settings.

Pay the provider, not us

Generation is billed by your AI provider at their rates. VisionMixer adds no markup — your $99 license already bought the studio around it.

A cap you control

An in-app monthly spending cap puts a ceiling on what a month of generating can cost. You set it, you change it, you're never surprised.

AI creation questions, answered

Which AI providers work with VisionMixer?

You bring your own keys: Claude (Anthropic) or Gemini for overlays and text, Gemini for images, Veo for video, and ElevenLabs and Gemini for voices, sound effects and music. Set them up once in Settings and every AI feature uses them.

What does generation cost?

Your AI provider bills you directly at their own rates — VisionMixer adds no markup. An in-app monthly spending cap you control puts a ceiling on what a month of generating can cost. The $99 one-time license covers the studio itself.

Do generated overlays need internet during the show?

No. Generating something new needs a connection to your AI provider, but the result is a self-contained file in your library. Once it's made, it runs in your show like any other overlay — no connection required.

Can I edit what it makes?

Yes. Generated overlays expose editable settings in the inspector — text, colors, timing — alongside their live controls. Tweak them directly, refine with another instruction, or regenerate from scratch.

Is my content shared with anyone?

Your prompts go only to the AI provider you chose, authenticated with your own key. Your data is between you and that provider, under their terms.

What does VisionMixer cost?

$99 once, with a 14-day free trial (trial output carries a small watermark). AI generation is billed separately by your provider, with no markup. Remote guests via Cloud Shows are an optional monthly add-on.