Open a VRM avatar’s Expressions section to toggle expressions, adjust weights, and change supported outfits.
Expressions
The Expressions section lists every morph target (blendshape) defined in the model, including VRM 1.0 presets and custom shapes; VRM 0.x preset names map automatically to their 1.0 equivalents. Click an expression button to activate or deactivate it, as with Live2D Expressions.
Face Tracking and Microphone Lip Sync drive these same blendshapes. Lip sync targets the standard vowel presets: aa, ih, ou, ee, and oh.
Each active expression displays a weight slider adjustable from 1% to 100%. Sliders appear only for active expressions; click the button again to toggle the expression off.
Models rigged with all 52 standard ARKit blendshapes support Perfect Sync, allowing tracking hardware to drive the facial mesh directly via raw ARKit channels without intermediate mapping. The Perfect Sync row in the info section indicates how many of the 52 channels are present.
Remember Expressions, enabled by default, preserves active expression states and weight values across scene changes and app restarts. Because VRM lacks .vtube.json files, settings are saved to a companion <model>.persona.json sidecar.
Outfits
Models containing the LAPLACE_outfit extension display an Outfit control group below their expressions. While standard VRM blendshapes handle mesh deformation, material colors, and UV offsets, they do not manage sub-mesh visibility. This extension provides visibility toggles and naming for modular clothing parts, ideal for avatars converted from platforms like VRChat. Persona handles outfit toggles through the expression pipeline, offering button toggles, transition sliders, persistent sidecar storage, and Plugin API control.
Facial expressions and outfit pieces are organized into distinct groups to prevent confusion between facial posing and clothing swaps.
Last updated on October 21, 2026