You don't need a $3,000 GPU to start making AI videos. Comfy Cloud gives you 5 free generations. I made a beginner guide showing exactly how to go from: Zero → Comfy Cloud → MiniMax H3 → reference image → AI video It takes about 5 minutes to follow. I documented every click so even if you've never touched ComfyUI before, you can do it. 📌 Bookmark this. 🔁 Repost it for someone getting started with AI video. Free tutorial in Comments

Testing lip syncing/ music video creation. Was expecting more from @openart_ai Song generation - Comfyui/ Yue2 Video generation/direction - OpenArt Director

Made a 2-minute short about a fox exploring a city reclaimed by nature. 🦊🌿 LTX 2.5 in ComfyUI generated the footage. Codex automated the 15-scene render, assembled the edit, balanced the audio, and added a title shot. “After We’re Gone” 👇

Testing out AI song generation + lip syncing. 🎵 Generated the song locally with YuE2 in ComfyUI 🎤 Used MinMax H3 for the lip sync The lip sync came out pretty decent — not perfect yet, but this was a fun experiment and I’m impressed with how far these workflows are getting. What do you think — convincing enough, or does the lip sync still give it away? 👀 If you’re experimenting with AI music or lip sync too, drop your workflow below. I want to see what everyone else is using.

The future of content creation is changing. I didn't open a single app. I spent 30 seconds writing one instruction to Hermes \u2014 and Hermes did the rest: it read the brief, wrote the prompt, spun up LTX-2.5 on my local GPU, rendered a 10-second photoreal shot, and posted it right here on X. The future of content creation isn't a better editor. It's no editor at all. You describe the outcome. The agent executes it end to end.

Meet NANNA — the fictional smart-home assistant that doesn’t take excuses. 😭 Snooze button? Curfew? Attitude? She’s got it handled. NANNA. Because someone has to raise your kids. My entry for the @OpenArt Ad Awards, created with Seedance 2.5. 🔊 #OpenArtAdAwards #AIVideo #Seedance25

I made a 30-second psychological murder-mystery thriller entirely with LTX 2.5 (open source). No actors. No camera. No Blender. No AI platform subscription costs. Generation time: under 5 minutes. This is where AI filmmaking is headed. Here’s the result 👇

Codex, Astra & Chill. 🎬 Built Crawl Spaces — my Backrooms-inspired found-footage experiment, except the maze is an endless attic. Codex built most of it for me. I focused on the concept, atmosphere, and direction while AI handled a huge part of the actual build. An experiment in what happens when AI becomes less of a tool and more of a production crew. CRAWL SPACES - Some places should stay unexplored. #AI #AIFilm #Codex #GenerativeAI #Backrooms

Sydney Sweeny new novig ad reaction! THIS IS INSANE!!!!

With ComfyUI set up, one scene prompt was enough. Codex used my PC to generate two shots of a boy cycling home, then joined them into one movie. 🤯 Created entirely in local ComfyUI: **two 10-second clips at 1024 × 576 (16:9)** #Codex #ComfyUI

Exploring longer movies with OpenArt's Director. Title: Crawl Spaces #openart #ai #liminalspaces

Attic Rooms - Compilation

Do you remember when you joined X? I do! #MyXAnniversary

Recreated the Juárez border crossing scene from Sicario — entirely in MiniMax H3 Turbo. Total gen time = 35 minutes. No reshoots. No VFX team. Just prompts. Comment "gimme" and I'll send you the full breakdown from visual transcription to prompt development👇

Attic rooms - found footage Made in Google Omni

Donald Trump just announced he's running for 2028? Wait what? And what's Ashton Kutcher doing?

found this camcorder tape in my parents' attic. we do not go up there anymore. 🎥 30 seconds. sound on. #FoundFootage #HorrorShort #Backrooms #Liminal #IndieHorror #Atticrooms

PDF tutorial > drive.google.com/file/d/15-s…