Innovate Futures @ Benji

(Demo) AI Singing Avatar Using All Local AI Models #aivideo #aianimation #aimusic

https://youtube.com/shorts/ISlrpv-iQRc?feature=share

Tutorial will be publish soon. stay tune :)

(Demo) Crafting New Workflow With Multi-AI

FramePack... it have it's own purpose for generate long video.

full video and demo video post later. :) This Gen Demo here >> https://youtube.com/shorts/bkNrkqUJA64

(ComfyUI Workflow) FramePack F1 + Flux ACE++ Start-End Frame - For Patreon Supporters

Tutorial for FramePack F1 : https://www.patreon.com/posts/128282706

Flux Ace++ 3In 1 Workflow : https://www.patreon.com/posts/workflow-flux-121648642

FramePack 2025-05-06 14:02:48 +0000 UTC View Post

FramePack F1 In WebUI and ComfyUI Locally - Update Model For Longer Better AI Video!

Tutorial Video : https://youtu.be/vEzRDZkZVgg

This video provides a comprehensive guide on using the newly updated FramePack F1 model, an advanced AI tool for creating high-quality, longer-length videos locally through both WebUI and ComfyUI. Viewers will learn how to download, install, and run the latest version of FramePack F1, co...

(ComfyUI Workflow) Wan 2.1 Fun Control Camera - Advance - 20250505

Related Tutorial : https://youtu.be/JiAxvau9qTE

Post: https://www.patreon.com/posts/128190474

This workflow integrated with Tile Lora for second sampling to help enhance the video generation.

Examples:

2025-05-05 13:00:18 +0000 UTC View Post

Wan 2.1 Fun Control Camera - Create Multi Camera-Motion With Low VRAM

Tutorial Video: https://youtu.be/JiAxvau9qTE

Create dynamic camera motion with Alibaba’s WAN 2.1 Fun 1.1 Control Camera! In this video, we test its ability to generate multi-motion sequences (pan, tilt, zoom) in AI videos, compare it to Recam Master, and reveal how it uses 1.3B/14B parameter models for lightweight performance. Wa...

Update - Wan VACE Workflows for WanVideoWrapper - 20250505

As recent update of WanVideoWrapper. Previous wrapper version built workflows might experience broken on ComfyUI, you might see the workflow layout broken or nodes disappear with you run.

I re-create Mask Inpaint and ControlNet Reference V2V workflow. The workflow are more simple and beginner friendly, less complex crazy node connections. ;)

For more information about VACE, please check out the previous video tutorials of VACE.

Previous Posts

Video Mask Inpaint : 2025-05-04 18:58:45 +0000 UTC View Post

(Demo Video)Just have fun using Wan AI, and I created this...

Although I am not a film maker, I am more on tech. But it's really fun to make a video like this. I know it's not prefect XD...

It's happy to see technology allow anyone to craft and create things from their idea.

Enjoy :)

AudioX AI - This AI Model Generate Anything-to-Audio Locally And Free!

Video : https://youtu.be/u9p-NceZzmI

AudioX AI - Diffusion Transformer for Anything-to-Audio Generation

Unlock the power of AudioX by HKUST—the AI tool that generates ANY sound or music from text, video, or images! In this video, we dive into its text-to-audio, video-to-music, and cinematic sound effect feature...

(ComfyUI Workflow) Fantasy Talking + Dia TTS + Flux - Multimodal AI Talking Avatar - Version 20250430

Tutorial Video : https://youtu.be/bSssQdqXy9A

Related Post: https://www.patreon.com/posts/127808092

Alibaba's Fantasy Talking represents a significant leap forward in talking avatar technology. When combined with Dia TTS for voice gen...

Fantasy Talking In ComfyUI Locally - Running Next Level Of AI Talking Avatar

Tutorial Video : https://youtu.be/bSssQdqXy9A

Discover Alibaba’s Fantasy Talking AI—the breakthrough in hyper-realistic talking avatars! In this video, we test its lip-syncing accuracy, body movements, and seamless audio integration using WAN 2.1 as the base model. Watch as we compare it to older AI avatars, optimize FPS ...

(Video Demo) Talking Avatar To The Next Level!

What do you think? Fantasy Talking + Dia :)

Update:
Workflow for freebies , model link, custom node links in this post:
https://www.patreon.com/posts/127808092?utm_source=youtube&utm_medium=video&utm_campaign=20250430

Additional content for Patreon Supporters: 2025-04-29 16:23:18 +0000 UTC View Post

Phantom Subject2Video Run Locally In ComfyUI - Better Then Wan 2.1 VACE?

Phantom ComfyUI Tutorial : https://youtu.be/2toaC3-yDcY

Unlock the power of ByteDance’s Phantom AI—the revolutionary subject-consistent video generation framework! In this video, we dive deep into Phantom’s ability to create seamless, multi-subject videos using WAN 2.1 as its base model. Watch as we test its performance...

ReCamMaster: Camera-Controlled - Version 20250427(Workflow Update)

Recent update with new Camera pose nodes.

Previous version : https://www.patreon.com/posts/recammaster-wan-126626847

(ComfyUI Workflow)Wan 2.1 First Last Frame To Video - Full Ver.20250424 - For Patreon Supporters

Tutorial Video : https://www.youtube.com/watch?v=7w3QR5sEhZw

Related post : https://www.patreon.com/posts/wan-2-1-first-to-127362075

Modify and additional features in this workflow.

Wan 2.1 First Last Frame To Video - Better Than Wan Fun Inp With 720P!

Video Tutorial : https://youtu.be/7w3QR5sEhZw


WAN 2.1's First & Last Frames To Video model—a revolutionary AI tool that transforms just two images into seamless video clips! In this video, we test its 14B-parameter, 720p-capable model in ComfyUI, comparing FP8 and FP16 versions for performance, and reveal how it o...

(ComfyUI Workflow) My Mod Skyreels V2 Diffusion Forcing Video Extend

Mentioned in this video tutorial : https://youtu.be/PBwDq_MDjKM
It's very easy to use, shouldn't have any problem to run this.

Other links and resources: https://www.patreon.com/posts/127221329

SkyReels-V2 - Another Long-Length Video AI Model But With Wan 2.1 ! (Tutorial)

https://youtu.be/PBwDq_MDjKM

Skyreel 2.0 is HERE—and it’s revolutionizing AI video generation with infinite-length videos! In this deep dive, we explore its groundbreaking diffusion forcing models, seamless video extensions, and how it outperforms traditional 5-second AI clips. Watch as we test 28-30 second ultra-smooth s...

(Video Demo) How To Use FramePack - 1

Using FramePack, T5TTS, LatentSync

FramePack Run In Gradio & ComfyUI - Generate Long Length image2Video AI Video - Installation Guide

Video: https://youtu.be/FE3beMmZObY?si=W-2OUh8vFad53Mmg

In this video, we explore two installation methods for generating long-length AI videos using the FramePack project. First, we demonstrate the easy standalone method with Gradio UI, showing how to set up and generate videos up to 60 seconds or more....

(Demo Video)Guys, please calm down when you see this...

https://www.youtube.com/shorts/ZIXqH-ScL-I

How crazy is that! Now, we can generate a long length AI Video (img2vid) , low VRam , locally.
Another big slap to those big tech! Testing now, and new tutorial next!

(Workflow)(Freebie)Hi-Dream ComfyUI Native Node Instruction And Workflow for txt2img With Upscaler

Tutorial Video : https://youtu.be/LyKhRvDWMRE

Files to Download

Download the text encoder files:

• clip_l_hidream.safetensors

• 2025-04-18 16:43:30 +0000 UTC View Post

Kling 2.0 Launched - Big Improvement For Img2Vid & New Video Editing Features

Video : https://youtu.be/eDo8RZA9qrc

In this video, we explore Kling AI 2.0’s groundbreaking features, including image-to-video conversions, dynamic character/object swapping, and cinematic fast-motion generation. Discover how it outperforms older versions (1.6) and rivals like WAN 2.1, delivering Hollywood-quality scenes, precis...

(Demo) Kling AI 2.0 Fast Motion Test

Hi @everyone As

Kling AI 2.0 released, I tested with some difference style of AI video using Image2Video generation.

On Youtube https://youtube.com/shorts/qlKpkqVQ1so

On X https://x.com/AIfutureBenji/status/1912799050526757325

ByteDance Uno In ComfyUI - The In-Context Image Generation Better Than Alibaba?

Tutorial Video : https://youtu.be/Qexw5nWokfI

Discover Bytedance's Uno Framework for AI-Powered In-Context Image Generation!

In this video, we dive into Uno, a cutting-edge framework built on Flux and LoRA models, designed to merge multiple image references into creative outputs. Learn how Uno handles style transfers,...

ReCamMaster Wan 2.1 In ComfyUI - Re-capture Video With Novel Camera Trajectories Using AI

Tutorial Video : https://youtu.be/uP0x0hinSTU

Transform your static videos into dynamic masterpieces using the groundbreaking ReCamMaster framework! In this video, discover how to inject AI-powered camera motions—like panning, tilting, zooming, and custom trajectories—into existing or AI-generated videos. Built to ...

HiDream-I1 Diffusion Model - The New Uncensored AI Image Model But Is This The Current Best?

Video: https://youtu.be/KRnJCLdgdRE

Discover the groundbreaking HiDream AI model, a 17-billion-parameter open-source text-to-image generation tool now available on Hugging Face! In this video, we explore its capabilities, compare it to Flux, and provide a step-by-step guide to running it locally using ComfyUI—even on consum...

Wan 2.1 VACE ControlNet Reference2Video And What Is the Difference With Wan Fun Control?

Video: https://youtu.be/NltDhu9LcAY

Additional Contents For Patreon Supporters : https://www.patreon.com/posts/126383882?utm_source=youtube&utm_medium=video&utm_campaign=2024041...

(ComfyUI Workflow) VACE Control Reference2video Ver. 20250410 - For Patreon Supporters

Tutorial : https://youtu.be/NltDhu9LcAY

Related Post: https://www.patreon.com/posts/126383882

The VACE ControlNet Reference2Video workflow stands out, enabling creators to generate dynamic, multi-subject videos guided by motion poses and style references. This blog post breaks down how to leverage this powerful method...

(Demo Video) VACE and Wan Fun Control - which one should you use?

I just did some test on bath VACE ControlNet v2v and Wan Fun Control.

Both method are using ControlNet as motion guide, but is Wan Fun Control are going to be obsolete since VACE released now? What do you think?