Hugging Apps

Tweeting the coolest apps and demos on @huggingface Spaces
curation by @multimodalart

[Hugging Apps @HuggingApps]

It’s a small model (0.5B) but good for its size class. Excited for this approach to scale https://t.co/c6CmVgRiUx

[Hugging Apps @HuggingApps]

WordVoice TTS ships what I've always wanted: a TTS system with per word control

you can let the system auto-pilot (based on CosyVoice3) or control every word with duration, loudness, pitch or tone

works with cloned or pre-set voices

▶️ on spaces https://t.co/ze8oJBqnbT

[Hugging Apps @HuggingApps]

UniSE is here to remove the background noise from any audio - finally open source caught up here

your voice memos recorded inside of a blender are salvageable now 🔊

try it for yourself on Spaces
▶️ https://t.co/RvMXqCzPmJ

[Hugging Apps @HuggingApps]

reshooting a video from a camera that was never there 📹

This LTX-2.3 IC-LoRA re-renders real footage from a NEW angle - same scene, same action, different camera. Pick "far to the left, higher, further" and it just… moves

Cooked by Cseti 🧑‍🍳 try it live on @huggingface:
▶️ https://t.co/PZ1TQQhvVX

[Hugging Apps @HuggingApps]

reshooting a video from a camera that was never there 📹

This LTX-2.3 IC-LoRA re-renders real footage from a NEW angle - same scene, same action, different camera. Pick "far to the left, higher, further" and it just… moves

Cooked by Cseti 🧑‍🍳 try it live on @huggingface:
▶️ https://t.co/PZ1TQQhvVX

[Hugging Apps @HuggingApps]

NVIDIA's Cosmos3 Edge is out! it watches videos streams & understands the mechanics/physics in them 🔥

it can reason in words, images, or next action prediction. physical AI reasoning, on the edge.

try it on @huggingface (or on your edge device)
▶️ https://t.co/MUrIG4dk2R

[Hugging Apps @HuggingApps]

Microsoft Asia just dropped Mage-Flow on Hugging Face

a smol 4B model for image generation and editing that matches much larger models in quality

and it's fast! 4 steps in < 1s at 1024x1024 (model goes up to 4K)

try out on spaces https://t.co/4hpT11IbD5

[Hugging Apps @HuggingApps]

@HuggingPapers Try out the demo for this model here https://t.co/caSfkSk9wz

[Hugging Apps @HuggingApps]

TimeLens2, the Ctrl+F for video model is here 🔥

ask when a certain event happens in a video, and get a time interval ⏱

Demo on @huggingface Spaces right now
▶️ https://t.co/RJLucNlzYj

[Hugging Apps @HuggingApps]

TimeLens2, the Ctrl+F for video model is here 🔥

ask when a certain event happens in a video, and get a time interval ⏱

Demo on @huggingface Spaces right now
▶️ https://t.co/RJLucNlzYj

[Hugging Apps @HuggingApps]

LTX-2.3 Foley LoRA dropped to add sound to silent videos 🔊

silent video in → sound effect out, generated from the pixels alone

press play on @huggingface spaces
▶️ https://t.co/sBJJig7GKM

[Hugging Apps @HuggingApps]

LTX-2.3 Clean Plate IC-LoRA by @ltx_io is fresh out of the oven - erasing people and cars from videos 👋

background, lighting and camera motion stay perfectly intact after the rapture 🎥

play with it on @huggingface spaces
▶️ https://t.co/mdewrj5PYS

[Hugging Apps @HuggingApps]

@drowsyleilei Thank you for open sourcing such a great model!

[Hugging Apps @HuggingApps]

@marzzelhere @huggingface It's a Krea 2 LoRA! Can run locally if you can run Krea 2 on your local device

[Hugging Apps @HuggingApps]

giving Jensen his leather jacket back with a single prompt 🧥🖤

this Krea 2 LoRA preserves identity and allows the model to follow edit instructions while keeping the person's likeness

cooked by conradlocke, live on @huggingface spaces

▶️ https://t.co/PadYZPI8BG

[Hugging Apps @HuggingApps]

NVIDIA Nemotron just dropped an audio-native model that hears the world, not just words 🔥

transcription, translation, sound recognition & audio Q&A, TTS and full speech-to-speech, hears, thinks, and talks back natively

open weights, 2B and 30B

▶️ https://t.co/jUv9HVSfr3

[Hugging Apps @HuggingApps]

@loktar00 It should run locally! https://t.co/TXZ0AnnMup

[Hugging Apps @HuggingApps]

literally insane how good Krea 2 is for outpainting when using this custom LoRA

yijunwang2 didn't train this LoRA, cooked it 🧑‍🍳🐐

it's now an app with a cool interactive canvas on @huggingface spaces

▶️ https://t.co/1tOASgGyuP