● Local-first studio
Tools
PIXELOKAY CREATOR LAB ✦ CHARACTER ANIMATION

Photo to Talking Character Studio

Turn a still portrait into a talking character. Add a voice, fine-tune lip movement and expressions, then download an animated video—all in your browser.

◉ Voice-reactive animation◎ Editable face points▤ Real video export✦ Face AI optional◆ Manual keyframes without voice

YOUR CREATIVE WORKSPACE

Talking character studio

Import. Animate. Fine-tune. Export.

Live character preview LIVE

Choose a portrait to begin.

720 × 720
☺

Your next character starts here

Open the sample or upload a front-facing portrait to start animating.

00:0000:00
Audio activity & manual keyframesUpload voice to see timing
◎ Face guides are editable. Select a landmark and click the preview to move it.
↧

Create your talking video

Video is recorded in real time from the animated canvas.

Clip selection

Import voice to select a clip. Short exports are ideal for creators.

Your browser chooses a supported WebM or MP4 codec. Voice is included when supplied and supported by the browser. Manual animation can export silently without any AI model or narration.

◈

Locally animated

Your images and audio remain in the browser. No external upload service. Optional face detection downloads its model only on request.

◎

You control the face

Correct the eye and mouth points and adjust the expressions.

▤

Real export

Download the recorded animation, with source audio where supported.

Ready. Open the demo to explore the editor.

WHAT YOU CAN MAKE

Talking portraits for lessons, stories and creators

Does lip-sync require AI?

No. Import narration for volume-driven mouth motion, or draw your own mouth keyframes without audio. These are stylized 2D effects, not phoneme-level alignment or photorealistic reenactment.

Does it work completely offline?

The editor, face-point controls and video exports work without a paid service. The optional face detector downloads free MediaPipe code and weights from external hosts when requested.

Can I use my own recorded voice?

Yes. Upload a supported audio file or record microphone audio when your browser permits it. The app processes audio on-device and lets you export a selected voice segment when supported.

Which export formats are supported?

Available formats depend on MediaRecorder support in your browser. WebM often works in Chromium. MP4 may be available elsewhere, but neither format is guaranteed on every device.

Animate a portrait with recorded audio

Set facial anchors on a portrait, add a short voice clip and review a stylized mouth animation driven by audio amplitude. The motion follows loudness, not spoken phonemes, so it does not reproduce precise lip-sync or verify who is speaking.

Example workflow

  1. Load a portrait you have permission to animate, then position the mouth and eye anchors manually.
  2. Import a voice clip or record through the microphone; optional Detect face can suggest landmarks.
  3. Preview the selected section, adjust captions or the background if needed, then export a still or voiced clip.

Authored example: Animate a mascot portrait to a short recorded welcome line. Check the mouth movement against the audio and label the result as stylized animation.

Limits and privacy: Audio is limited to 60 seconds. Microphone recording requires browser permission and a secure context. Detect face can request optional MediaPipe code, WASM and face-landmarker model assets from external hosts; delivery and inference were not checked in this source review. Settings JSON excludes the portrait and audio. The output is not authenticated speech.

Related tools