Vois
Back to BlogProduct Updates

Introducing Omni and Vois Pro: 600+ languages and Voice Design on your machine

Vois TeamVois Team
June 9, 2026
5 min read

TLDR:Vois 2 introduces Omni, an on-device model for 600+ languages and Voice Design. Pro adds Omni to the core local studio so teams can localize and cast new voices without per-character billing.

A localization project usually creates two problems at once: you need the right language coverage, and you need a voice that still sounds like it belongs to your brand. Vois Pro brings both decisions into the same local production workflow with Omni, so you can write, cast, review, and export without sending the script to a per-character service.

Omni adds more than 600 languages and Voice Design to Vois. The core studio still handles everyday production with fast, expressive, and multilingual engines. Omni is the option for a broader language brief or for a voice you cannot find in an existing library.

Launch illustration for the Vois Omni and Pro release

What Omni adds to Vois

Omni is an on-device voice model available with Vois Pro. It does two jobs that change the shape of a multilingual production workflow.

First, it expands language coverage to more than 600 languages. The multilingual engine in the core studio covers 23 languages, which is a good fit for many projects. Omni gives localization teams and creators a wider set of languages without moving the rest of their production out of Vois.

Second, Omni powers Voice Design. Instead of starting with a recording, you select the attributes you need for a new voice, preview how it performs with your own text, and save the result for the project. That gives you another casting route when a library voice is close but not right.

The local part remains the point. Your script stays on your device while you iterate on language, delivery, and voice selection. You can retry a line because it needs a better pause, not because you are trying to conserve a character allowance.

How 600+ languages change localization work

Language count does not make a release ready by itself. A useful localization workflow starts with a script adapted by someone who understands the target audience, then tests a representative passage for names, tone, punctuation, and pacing.

In Vois, that review happens where the production happens. Create the target-language script, choose Omni, audition a voice against the passage, and collect feedback before generating the complete version. If a product name or technical term needs special treatment, correct it in the script or pronunciation workflow and listen again.

Globe illustration for broad language coverage

That approach gives a small team a practical path to consistent regional versions. A solo creator can keep a familiar tone across several languages. A training team can review each local version before it reaches employees. The work is still editorial, but the voice production no longer requires a separate vendor and a separate billing meter for every revision.

Use Voice Design when a library voice is not the answer

Voice Design lets you create a new Omni voice from selected attributes. You decide the qualities that matter for the role, preview the result with the script, and save the approved voice for reuse.

That is different from voice cloning. Cloning starts with a real person's audio and requires their clear permission. Voice Design begins with selected attributes for a new voice. Both belong in the Vois library once approved, but they solve different casting problems.

Use a library voice when it fits. Use a clone only when you have the right consent and need that specific person's voice. Use Voice Design when the character or narrator does not already exist. The decision is easier when all three routes lead back to the same script editor, speaker assignments, timeline, mastering, and export tools.

Voice Design workflow illustration

Test Omni on the hardware you will use

Performance depends on the computer and the material you are generating. Apple Silicon uses GPU acceleration for faster Omni generation. Other supported hardware can behave differently, so a production estimate based on somebody else's machine is not a useful promise.

After Pro access is active, run a representative script through Omni and listen to the quality and pacing before you plan a large localization batch. Test the names and sentences that tend to fail, not only a clean demo paragraph.

Hardware testing before a voice production project

Choose the plan around the work you need to ship

Plan Production fit
Subscriber Core local studio, 100+ voices, voice cloning, and fast, expressive, and multilingual engines
Pro Everything in Subscriber, plus Omni, 600+ languages, and Voice Design

Subscriber is a strong fit when the core language set and voice library cover your work. Pro is the right fit when you need Omni's broader language range or need to design a new voice for the project. Both keep the same local studio workflow: write the script, cast the speakers, generate, review, master, and export.

See current plan details when you are comparing options. Do the comparison around the languages, voice source, and review volume your team actually needs, not around a single showcase clip.

Start with a production test, not a promise

Omni widens what a local voice studio can do, but the workflow remains deliberately practical. Adapt the script. Test the target language. Audition the voice. Generate a review candidate. Fix what the listener hears. Then export the approved version.

Explore Voice Design and the Vois library, compare current Pro details, and Get started. Once Pro access is active, test Omni with your own representative script.

The Vois Team

Frequently Asked Questions

What is the Omni model in Vois?

Omni is Vois's on-device model for more than 600 languages and Voice Design. It is available with the Pro plan.

Does Omni need a GPU?

Omni is a Pro feature. Performance varies by machine, and Apple Silicon uses GPU acceleration for faster generation. Test it on the machine you will use for production after Pro access is active.

What is the difference between the Pro and Subscriber plans?

Subscriber includes the core local production workflow, 100+ voices, voice cloning, and the fast, expressive, and multilingual engines. Pro adds Omni, 600+ languages, and Voice Design. See the pricing page for current plan details.

How many languages does Omni support?

Omni supports more than 600 languages. The multilingual engine in the core studio covers 23 languages, while Omni extends that range for broader localization work.

Can I try Omni before buying?

Omni is available with Pro. Review the current plan details, then test a representative script on the production machine after Pro access is active.

AnnouncementReleaseAi VoicesMultilingualProduction
Share:
Vois Team

Written by

Vois Team

Product Team

The team behind Vois, building the future of AI voice production.