Leonardo Fries

Video Editor / 3D / Graphic Design / AI Enthusiast
I'm active, just not posting much

[Leonardo Fries @oLeonardoFrito]

Tired of paying for expensive APIs?

Meet Director’s Cut. Create characters, props and locations, then bring them together in video—all locally with ComfyUI.

Open source. Your cast. Your direction.

https://t.co/siG6UKr0nq

#OpenSourceHiggsfield

[Leonardo Fries @oLeonardoFrito]

$NVDA is down 5% again?
When do the gpu prices start dropping too?

[Leonardo Fries @oLeonardoFrito]

@ComfyDevs @noma_door The competition was good.
Loved partaking on the challenge anyways.

[Leonardo Fries @oLeonardoFrito]

Having fun with Second Take.

Update coming later this week

[Leonardo Fries @oLeonardoFrito]

Your turn, OpenAI

[Claude @claudeai]

Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges.

They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.

[Leonardo Fries @oLeonardoFrito]

@BennyDaBall_OG Managed to go as low as 150s using 6 steps, comfy kitchen and the 2-passes self lift sampler (3 steps at 0.5mp, and 3 steps upscaled to 1mp)

[Leonardo Fries @oLeonardoFrito]

Thanks to @BennyDaBall_OG HyperFlow H3 Flash INT8 ComfyUI weights, I've got another fast model to run and use on Director's Cut.

God, I love open source models.

5s @ 1MP - 8 steps with Comfy Kitchen: 236s
RTX 3080Ti 12GB + 32GB RAM 🔥

[BennyDaBall @BennyDaBall_OG]

HyperFlow H3 Flash shipped INT8 weights with no way to run them in ComfyUI. So I built one. 🔥

The fastest H3 setup I've ever run at this resolution, and a real speedup: their own INT8 weights, copied bit-exact, running on ComfyUI's native kernels with one loader node.

One RTX 5090, 1344x768, 5 s with audio:
46.1 s, vs 66.0 s for the official GGUF
About 38 s with sparse attention
3x faster than stock H3

All clips AI-generated locally on a single 5090.

Links, nodes and weights below. 👇

[Leonardo Fries @oLeonardoFrito]

@BennyDaBall_OG Video Rebirth also released their own INT8 Comfy version.

Same prompt, seed and settings: 276s @ 1MP 5s

[Leonardo Fries @oLeonardoFrito]

#OpenSourceHiggsfield

[Leonardo Fries @oLeonardoFrito]

Forget Genjutsu, try Second Take inside Director's Cut.

Multiple characters, multiple cuts, generate and iterate infinitely.

Create your own viral video today with Director's Cut
https://t.co/siG6UKr0nq

[Leonardo Fries @oLeonardoFrito]

Director’s Cut rc.8: Redesigned, streamlined UI.

Second Take is our take on Genjutsu:

Multiple subjects, multiple cuts.

Built-in timeline, MP4/DaVinci export.

100% local, 100% open source.

https://t.co/siG6UKqsxS

[Leonardo Fries @oLeonardoFrito]

Tired of paying for expensive APIs?

Meet Director’s Cut. Create characters, props and locations, then bring them together in video—all locally with ComfyUI.

Open source. Your cast. Your direction.

https://t.co/siG6UKr0nq

#OpenSourceHiggsfield

[Leonardo Fries @oLeonardoFrito]

Major update incomign in the next 30 minutes or so.

Complete remodel of the app + character swap with Multiple Subject support and Multiple Cuts! That's right, characters will stay consistent across multiple cuts!

[Leonardo Fries @oLeonardoFrito]

Tired of paying for expensive APIs?

Meet Director’s Cut. Create characters, props and locations, then bring them together in video—all locally with ComfyUI.

Open source. Your cast. Your direction.

https://t.co/siG6UKr0nq

#OpenSourceHiggsfield

[Leonardo Fries @oLeonardoFrito]

O designer que fez isso tem um amplo portfólio de criação de logos para mockups de empresas de delivery

[bradesco @Bradesco]

Aguarde, tem novidade chegando. ⏳

#NovaFotoDePerfil

[Rikiya Takehi @rikiyatakehi]

We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.

We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: https://t.co/bOWCcFhCVq

[Leonardo Fries @oLeonardoFrito]

If it were not for the 5-hour limit. I almost have Higgsfield "Genjutsu" fully figured out using Minamax H3 and some other custom nodes. I'll try and drop a video on it today.

#OpenSourceHiggsfield

[Leonardo Fries @oLeonardoFrito]

Damn you 5h limit

[Leonardo Fries @oLeonardoFrito]

@itsnex1s Hey, that's me right there!
Nice resource anyways, I like it

[Leonardo Fries @oLeonardoFrito]

👀

[Leonardo Fries @oLeonardoFrito]

Making updates to Director's Cut, who knows, maybe I'll do a OpenJutsu to run on low vram systems like mine and a Influencer Studio

[ComfyUI @ComfyUI]

The community support has been incredible, and we’ve been blown away by the creativity and amazing apps being submitted.

We’re extending the submission deadline to November 3rd at 9 AM PT to give everyone a little more time to participate.

We value this community and love seeing what you’re building.

If we win, we all win. 🫡