NoLang Adds Real Avatars and Voice Clones From One Photo and One Minute of Audio

Mavericks, Inc., the Japanese company behind the generative AI product NoLang, has released a major update to its AI video generation service NoLang that lets users build an avatar and a voice from their own photos and audio. On the Standard and Premium plans, a single photo and one minute of audio are now enough to create a digital stand-in that can appear in videos. The Business plan goes further, adding lifelike “real avatars” and high-quality voice clones generated from existing audio files. Individual creators gain a way to make their output unmistakably their own, while companies can fully automate videos featuring their executives or their characters at a far lower cost.

A long-awaited avatar and voice creation feature

NoLang is Japan’s first AI video generation service: users enter text, a PDF or PPTX document, a website URL, or an image, audio or video file, and a video is generated in as little as a few seconds. Since its release in July 2024 the service has passed 150,000 registered users and is now used by more than 60 companies across Japan (as of November 2025). This update moves the service beyond reusing existing material. Users can now bring in their own photo, an image-based IP character owned by their company, or a recording of their own voice as source material. The workflow stays as simple as before, but the finished video now carries the speaker’s identity or the company’s brand image directly.

For Standard and Premium users: a digital stand-in from one photo and one minute of audio

This feature is available on the NoLang Standard and Premium plans. Creators and individuals publishing on their own can instantly produce videos with a face and a voice of their own, without special equipment or skills.

Create an avatar that looks just like you from a single photo

Upload one photo and your own illustration or original character appears in the video as a 2D avatar. YouTubers, streamers and influencers who prefer not to show their face can animate a character they drew themselves, protecting their privacy while still speaking to viewers with plenty of expression. The same applies elsewhere: an avatar delivering the explanation in an internal training video adds visual movement that helps employees stay focused, and reviews from users who do not want to appear on camera can be turned into UGC-style videos.

Create a voice that sounds like yours from just one minute of recording

The Create voice feature needs only one minute of audio, recorded on the spot, to produce a voice clone that reproduces the character of your voice with high fidelity. Narration no longer has to be recorded for every video, and the finished video speaks in your own voice rather than a generic AI voice, which helps build engagement with your audience.

For business customers: real avatars and high-quality clones for business video

The Business plan includes stronger versions of both the avatar and the voice creation features. They address two problems that hold back corporate video: the difficulty of getting executives and managers in front of a camera, and the long production schedules and swelling costs that come with high-quality video.

Create a lifelike animated real avatar from a single photo

On the Business plan, a single photo produces a “real avatar” that closely resembles the person in it. Lip-sync technology reproduces mouth movements, so the avatar looks as though the person is genuinely speaking. Message videos from a busy CEO and internal newsletters can be published on time from text input alone, with no shooting time booked for the executive. The same technique makes training videos feel as though the instructor is speaking to the viewer in person, which supports retention of what employees learn.

Turn a static image of your company’s anime character into a moving avatar

The high-quality anime character creation feature animates an image-based official mascot or an IP character a company owns, at high resolution and without breaking its art style. A character that until now existed only as a still image can act as a spokesperson and help grow brand awareness on social media. A local government mascot can introduce its town in a video, and eye-catching vertical character ads can be produced for digital signage and similar placements.

Voice clone generation from existing audio files

Voice clones can now also be generated from existing audio files such as mp3. When the executive or manager who should appear in the video cannot be scheduled, a clone can be built from high-quality audio recorded in a studio in the past or from a recording of a talk they gave. Entering a script is then enough to produce convincing training videos and manuals in that person’s own voice, at volume, without them being present.

Looking ahead

Mavericks, Inc. intends NoLang to be more than an efficiency tool, and is developing it into a next-generation infrastructure for video communication that maximizes both individual expression and the value of corporate assets. The company will strengthen customized avatar and voice development for the specific needs of individual businesses, and will continue to expand the feature set and advance the underlying technology, moving from a world where anyone can make a video to one where anyone can make a video that sounds like themselves and conveys what they mean.