SOURCE-REVIEWED MODEL PROFILE
AURA-1.3B
Audio-driven streaming-avatar student; non-commercial research weights, not a consumer-GPU speed guarantee.

Metadata reviewed 2026-11-20 · Original creator: Kyutai
What it does
- Portrait plus speech input
- Four denoising steps per chunk
- Merged and LoRA-only research artifacts
Local fit
Publisher:576p optimized20.7fps/H100;referencecompiled about16fps/H100. Not locally tested.. Separate text/audio/video components required. A small model label is not total memory.
Video streaming uses a rolling cache;agent adapter context depends on the base and runtime. No independent long-context test.
Evidence and limits
Source reviewed;no local inference,hardware fit or benchmark test.
Publisher reference workflow;not independently tested
License and source
Card explicitly declares CC BY-NC4.0; no standalone LICENSE file verified. Dependencies and code retain separate terms.
Read the original model card ↗
Reviewed source revision: 53558287fd7c62d33fc4217f5f4e7b8d2d729526