Mavericks, Inc., the Japanese company behind the AI video generation service NoLang, which has passed 150,000 registered users, has launched an avatar generation feature that turns a single uploaded photo into a video avatar that speaks with emotion. The feature builds an image avatar from a selfie taken on the spot, an uploaded character illustration, or an older photo, and that avatar can express emotions such as joy and anger and move its mouth in sync with the audio (lip-sync). Supported file formats are heic, jpg, jpeg, png, webp, avif, tif and tiff. Individual creators can now pursue a more original look while publishing without showing their face, and companies can put illustrations of their own characters to work as mascot characters in video. Neither specialist knowledge nor extra cost is required, so anyone can take part as a creator.
One image is all it takes to build an avatar, so anyone can make original video
NoLang is Japan’s first AI video generation service that produces a video in as little as a few seconds from text, PDF or PPTX documents, a website URL, or image, audio and video files. Since its release in July 2024, registered users have passed 150,000, and more than 60 companies have adopted it, with wide use across Japanese businesses as of November 2025. NoLang has long offered over 100 stock avatars as well as real avatars that closely resemble live-action footage, but animating a user’s own original character still required specialist production skills such as Live2D, or an outside contractor.
With this update, users can upload a still image they already have, whether an illustration or a photo, and immediately have it appear in a video as a moving avatar, called an image avatar. Creators can now build videos that reflect their own identity intuitively, expressing originality more easily and more deeply.
Three characteristics of the image avatar generation feature
Emotion and movement generated automatically from a single photo to create an avatar
The defining characteristic of the feature is that AI can build an avatar that reflects a character’s individuality and actually moves in the video, all from one still image. When an image is uploaded, the AI creates a total of eight avatar styles, and the user picks one of them. The AI then generates emotional versions of that avatar, such as joy, anger and sadness, and selecting up to six of those emotions completes the image avatar. The avatar does more than move its mouth in time with the audio in the video (lip-sync): it also sways its body to match the emotion or the idle state, bringing a still image to life.
Pair it with a custom voice to create an avatar that sounds just like you
Combined with the voice clone feature of NoLang, which creates a voice that reproduces your own from a one-minute recording, users can create a fully original avatar customized in voice as well as appearance. That gives branding and character expression a consistent feel both visually and audibly.
From people to animals to IP characters, any original character can appear in a video
The source image does not have to be a photo of a real person. Uploading a character image is enough to create an avatar of a company’s own character on the spot. Beyond human avatars, any material, including animals and in-house mascot characters (IP), can be used as a moving avatar. Companies can therefore use assets they already own to produce videos with an originality that sets them apart from competitors.
Use cases: from personal publishing to making use of corporate assets
The feature helps individual creators make their videos more expressive and their publishing more original, and it helps companies cut costs and put dormant assets to good use.
[Influencers and individual creators] Stronger personalized publishing
For creators who want to publish vlogs, gameplay commentary or explainer videos without showing their face, the feature is a powerful tool. Turning a caricature of yourself or an original character into an avatar preserves anonymity while leaving viewers with a strong sense of familiarity and personality. Staging a vlog so that your own character appears to be speaking raises that sense of familiarity, and adding avatar movement to an explainer video helps viewers stay focused, both of which contribute to higher engagement and retention.
[Communications and PR] Differentiated promotion using illustration assets already held in-house
Corporate mascot characters and illustration assets that sit unused in a corner of a website or as a spot illustration in a flyer can be revived as new assets through NoLang. With this feature, those still photos and images can appear in video as moving characters at no additional cost. Putting dormant assets into social media ads and company introduction videos delivers memorable promotion that stands apart from competitors while keeping costs low.
[Social media marketing and advertising] Turning UGC and customer testimonials into video
When customer reviews and interviews (UGC) are turned into video content, privacy concerns can make it difficult for people to appear on camera. Even then, having an approachable avatar speak on their behalf conveys warmth and credibility that text alone struggles to carry. Presenting the material as video also brings text and image reviews, which are easily overlooked, to a wider audience, which in turn builds product awareness and deeper interest.
Outlook
Mavericks, Inc. will continue to improve the expressive range of the avatars available in NoLang and to expand the supported formats. The company aims to remove every technical hurdle in video production and to become the infrastructure of an era in which anyone, individual or organization, can take part as a creator.
