Paper 2604.07823
LPM 1.0: Video-based Character Performance Model
- Published
- Apr 2026
- Research lab
- Independent
- Citations
- 5
- GitHub
- 361 stars
01 In brief
Summary
LPM 1.0 is a video-based character performance model that generates identity-consistent conversational videos in real time.
It addresses the 'performance trilemma'—the challenge of achieving high expressiveness, real-time inference, and long-horizon identity stability simultaneously.
The system comprises a 17B-parameter Diffusion Transformer (Base LPM) trained on a large-scale multimodal dataset, and a distilled causal streaming generator (Online LPM) for low-latency, infinite-length interaction.
The dataset includes speaking and listening audio-video pairs, identity-aware multi-reference images, and detailed captions.
Base LPM uses interleaved audio cross-attention and multi-reference conditioning for controllable generation.
Online LPM uses a backbone-refiner architecture with a four-stage distillation curriculum.
The authors introduce LPM-Bench, a benchmark with 1,000 test cases across five scenarios, and report state-of-the-art results in human evaluations, with Base LPM preferred over Kling-Avatar-2 and OmniHuman-1.5, and Online LPM preferred over LiveAvatar and SoulX.
The system is designed for applications like conversational agents, live streaming, and game NPCs, with safety measures including watermarking and consent protocols.
02 From the paper
Abstract
Performance, the externalization of intent, emotion, and personality through visual, vocal, and temporal behavior, is what makes a character alive. Learning such performance from video is a promising alternative to traditional 3D pipelines. However, existing video models struggle to jointly achieve high expressiveness, real-time inference, and long-horizon identity stability, a tension we call the performance trilemma. Conversation is the most comprehensive performance scenario, as characters simultaneously speak, listen, react, and emote while maintaining identity over time. To address this, we present LPM 1.0 (Large Performance Model), focusing on single-person full-duplex audio-visual conversational performance. Concretely, we build a multimodal human-centric dataset through strict filtering, speaking-listening audio-video pairing, performance understanding, and identity-aware multi-reference extraction; train a 17B-parameter Diffusion Transformer (Base LPM) for highly controllable, identity-consistent performance through multimodal conditioning; and distill it into a causal streaming generator (Online LPM) for low-latency, infinite-length interaction. At inference, given a character image with identity-aware references, LPM 1.0 generates listening videos from user audio and speaking videos from synthesized audio, with text prompts for motion control, all at real-time speed with identity-stable, infinite-length generation. LPM 1.0 thus serves as a visual engine for conversational agents, live streaming characters, and game NPCs. To systematically evaluate this setting, we propose LPM-Bench, the first benchmark for interactive character performance. LPM 1.0 achieves state-of-the-art results across all evaluated dimensions while maintaining real-time inference.