LOADING DATASHEET
LOADING DATASHEET
by Tianweiy
RANKED #62 OF 101 IMAGES · OVERALL #296 OF 493 RANKED · VIBE SCORE 5.9 · 96 VOICES ·
COPY BADGE pastes a README snippet with the rank SVG.
DMD2 on Hugging Face (text to image). 28,860 downloads. Open weights for local or hosted use.
25 mentions
7 mentions
26 weeks · 37 voices in this window
HOLDING UP
Recent vibe is steady. No clear fade in the crowd.
Open weights. You can use a hosted service, or download it and run it yourself, free.
Open weights: download DMD2 once and it's yours. No subscription, no rate limits, works offline.
2.6B parameters (SDXL distilled UNet), FP16 checkpoint is about 5.2 GB
NVIDIA GeForce RTX 3060 12GB
12 GB VRAM
~$280 USEDFP16 4-step generation at 1024x1024, ~1.2 s/image
NVIDIA GeForce RTX 4070 12GB
12 GB VRAM
~$520 NEWFP16 4-step generation at 1024x1024 with xFormers, ~0.6 s/image
NVIDIA GeForce RTX 4090 24GB
24 GB VRAM
~$1800 NEWFP16 4-step or 1-step batch generation at 1024x1024, ~0.2 s/image
The three cards above are researched picks. Search the machine you actually have — we only say yes if it should reply at a usable speed, not just load the weights and crawl.
HARDWARE PICKS & PRICES ARE RESEARCHED FROM THE LIVE WEB AND REFRESHED AUTOMATICALLY. TREAT THEM AS BALLPARK, NOT GOSPEL.
Everyday jobs on each of those rigs: how long DMD2 takes, and how good the crowd says it is at that kind of work.
| THE JOB | BARE MINIMUMNVIDIA GeForce RTX 3060 12GB | SWEET SPOTNVIDIA GeForce RTX 4070 12GB | FULL POWERNVIDIA GeForce RTX 4090 24GB |
|---|---|---|---|
| Make an imageone picture at normal settings | ~1 S★★★★★ | UNDER A SECOND★★★★★ | UNDER A SECOND★★★★★ |
STARS ARE THE COMMUNITY SCORE FOR THAT KIND OF WORK, DOCKED FOR HOW SQUASHED THE WEIGHTS GET AT EACH TIER. TIMES ARE COMPUTED FROM RESEARCHED SPEEDS. BALLPARK, NOT GOSPEL.
Not enough discussion about DMD2 yet to call it. We found 96 posts but people did not say much either way.
63 POSTS · 33 COMMENTS · LEMMY · REDDIT · GITHUB
Other models the crowd has fully reviewed, starting with images models like this one.
Nvidia releases Qwen-Image-Flash
DMD2 is fast
TwinFlow - Qwen Image with 2 steps.
This is Playground V2.5 with a 20% DMD2 Refiner (14 pictures)
SeFi-Image: A Text-to-Image Foundation Model with Semantic-First Diffusion
Small dusty/dirty images with SD3.5 Large, detailed/cleaned with DMD2
DMD2 and euler_dy/bong_tangent, these images, nice revisit to SDXL. WF in comments.
Anima-Turbo Coming Soon
[Release] ComfyUI-AutoGuidance — “guide the model with a bad version of itself” (Karras et al. 2024)
8-step FLUX.2-dev DMD2 distillation
4-step SDXL DMD2, colab, images (not upscaled) Tuto and link:
Upscaling old low resolution screenshots from Dune 2 with SDXL