Live Caption Studio

Computer and audio requirements

Plan for an NVIDIA RTX GPU on Windows or Linux, or a Mac with Apple Silicon, and a clean speech feed from your sound console. These hardware tiers are planned, confirmed by our evaluation before the beta.

Planned hardware tiers

Planned requirements, confirmed by our evaluation before the beta.

Minimum

Captions only

  • NVIDIA RTX 3060 12 GB or RTX 4060 8 GB, with 16 GB of memory
  • A Mac with M2, M3 or M4 and 16 GB of memory

Recommended

Captions and translation

  • NVIDIA RTX 4070 Ti Super, 4080 or 5070 Ti 16 GB, with 32 GB of memory
  • A Mac mini with M4 Pro and 24 GB of memory or more

Not supported

In version 1.0

  • Computers without a graphics card
  • AMD and Intel graphics
  • Intel Macs

Choose a tier for the job

The minimum tier is the planning reference for captions only with a small model. The recommended tier is for captions with translation. GPU memory and system memory are separate requirements on the NVIDIA rigs.

These examples come from the product requirements, not a completed hardware benchmark. A matching model name is not a measured performance guarantee. We will confirm the tiers before the beta; no automated compatibility verdict is offered before that evaluation.

Hardware outside the version 1.0 plan

CPU-only computers, AMD or Intel GPUs and Intel Macs are not supported in the version 1.0 plan. A machine running Windows, Linux or macOS is not sufficient by itself: the graphics hardware matters.

The LCS Box is a planned hardware-plus-license option for venues that want a dedicated unit. Its final specification and price are not confirmed. Tell us what you need on the Box interest list.

Don’t want to build a PC? → LCS Box

Audio is part of the requirement

Live Caption Studio needs a clean speech feed: the microphone routed directly from your sound console (a post-fader aux or matrix), not the room or the music mix. We make no accuracy promise for speech over a music bed.

Use a dedicated bus for speech microphones, excluding playback, instruments, ambient microphones and effects returns. A post-fader send follows the channel fader and mute, which helps keep private stage conversations out of the caption feed. Check the routing at the console before a show.

For best results

  1. Keep music out of the dedicated speech feed; check the routing before the show.
  2. Make sure speech is loud enough. Set gain so normal speech sits around the console’s nominal level, without clipping.
  3. Keep the microphone close and steady. Watch for speakers who lower their voice or step away.
  4. Use one voice at a time: overlapping speakers reduce recognition accuracy. Interpreters should wait for a pause; on panels, keep one microphone open at a time where possible.

Read the clean speech feed guide

Connections to check before the beta

Console USB audio or an audio interface can carry the dedicated speech bus to the computer. Choose the intended device and channel; confirm the speech microphones reach it and music does not.

The planned ATEM linear key output needs two available HDMI outputs: one for fill and one for key. Reserve those outputs when planning the computer and the rest of the booth displays. Phone viewing and browser overlays use different output paths.

Activation and network planning

The application is built to run without internet at show time. Free needs no activation and works offline indefinitely. One-time licenses work offline indefinitely on the purchased major version after a single activation (online, or by file from another device). Subscriptions renew once a month, online or by license file, with a 7-day grace; without renewal, the installation then continues as Free. Optional YouTube captions send caption text, never audio, to YouTube and need that online destination.

In the beta form, tell us your operating system, exact GPU or Mac chip, memory and audio route. Include whether you need translation and ATEM fill/key so we can understand the complete setup.