HUST + VIVO AI Lab · 2026-06-18 · notable
Moebius — 0.22B image inpainting matches FLUX.1-Fill-Dev's 11.9B
Moebius is a 0.22B image inpainting model from HUST and VIVO AI Lab that matches the 11.9B FLUX.1-Fill-Dev across six benchmarks while running over 15x faster, with code and weights now on GitHub.
A 226M-parameter inpainting model that keeps up with 11.9B systems and runs 15x faster.
Key specs
| Parameters | 0.22B |
|---|---|
| Speedup | 15x |
| Step latency | 26.01 ms |
| Compared to | FLUX.1-Fill-Dev 11.9B |
What is it?
Moebius is a lightweight image inpainting framework from HUST and VIVO AI Lab. The 0.22B model is under 2% the size of FLUX.1-Fill-Dev (11.9B) and matches or beats it on six benchmarks across Places2, CelebA-HQ, and FFHQ. Moebius's training and inference code plus model weights shipped on June 18, 2026.
How does it work?
Moebius introduces a Local-lambda Mix Interaction block to fight the representation bottleneck of extreme compression, paired with adaptive multi-granularity distillation that transfers knowledge from a 10B-class teacher. The architecture targets local context at multiple scales, which lets a tiny network preserve detail in masked regions. Per-step inference runs in 26.01 ms.
Why does it matter?
Moebius shows that high-quality inpainting does not need a billion-parameter generator. Teams that could not afford to deploy FLUX.1-Fill-Dev now have an Apache-2.0 model 50x smaller that runs on commodity GPUs. The 15x speedup also opens up interactive editing flows where each user click triggers a new inpaint.
Who is it for?
ML researchers building image editing tools, product teams adding inpainting to apps, students studying model distillation.
Try it
git clone https://github.com/hustvl/Moebius