For studios

Translated captions across every machine you run.

XNN Translate runs on each machine's own GPU, so studios can caption several shows at once — up to two devices on Insider and three on Sponsor, all on one account. Plan a larger rollout with us anytime.

  • Multi-device on one account
  • On-device
  • Talk to us

Studio workflow

Run captions on several machines from one account.

Studios and multi-creator setups run more than one machine, each captioning locally on its own GPU. Everything runs on-device — there is no cloud caption service to manage.

Multiple devices on one account

Add machines to a single account as you grow — Insider covers two devices and Sponsor three, each running the AI locally.

  • Insider 2 devices
  • Sponsor 3 devices
  • One account
For studios

Consistent per-machine setup

Give every machine the same install, OBS output, and caption settings so shows look the same wherever they run.

  • Same install
  • Same OBS output
  • Same caption look

One place for help

Keep setup and caption questions in one account, and reach out to plan a larger multi-device rollout.

  • Account support
  • Setup help
  • Plan a rollout — talk to us
Get help

Operating model

Consistent setup across machines without slowing individual shows.

Studio content connects one account, its devices, and a consistent per-machine setup to the same caption workflow creators already use.

Add your devices

Activate each machine on the same account so every show can start captions under the right plan.

  • One account
  • Add devices
  • Plan state
For studios

Standardize the look

Keep install, OBS output, and caption styling consistent across recurring shows, hosts, and audiences.

  • Install
  • OBS output
  • Styling

Plan a rollout

For a larger multi-device or multi-creator rollout, talk to us so setup and account questions stay in one place.

  • Multi-device rollout
  • Setup help
  • Talk to us
Get help

Ready for translated captions?

Choose the next product action that matches your workflow.