SD | AI Animation Storyteller (@SDXLHQ) header
SD | AI Animation Storyteller (@SDXLHQ) profile picture

AI animator turning scripts into cinematic shorts 🎬 🇨🇦 | Drop a follow for Kling/Niji/Sora secrets most creators 🚀 #AIArt

Ottawa, Canada
Joined November 2022
🚨 @OiiOii_AI: Turned my story idea into a FULL anime short film in minutes! I fed @OiiOii_AI one prompt + ref image... and its agents built the entire film. No manual grinding! 📷 Agents handled script, characters, storyboards, music, videos – no manual grind. Game-changer for creators! Can Improve: -> Bugs in complex scenes (e.g., character consistency glitches) -> Deeper customization (e.g., palette overrides) -> Some clips have background music, but I managed to override the prompt and remove it. Workflow: Art Director agent analyzes/sets fields, hands off to Scriptwriter agent (script/cost), to Character Designer agent (images/concepts), to Storyboard Artist (music/scenes/animations/videos). Agentic chain: Updates, generates, "invites" - like a virtual team! So the question is, is it perfect? Almost there, it's really commendable how far it has come to solve some problems creators face. Full points for that. I'm sure with time it will be better. I loved the experience and wish the best. Detailed workflow below 👇💥 #aivideos #AgenticAI #Oiioii #AIart
24
30
5
331
25,473
Just tried Seedance 2.5 on @dreamina_ai These 30-second outputs are looking really promising 🔥 #DreaminaCPP What do you think of this one?
🤖 Made with AI
11
346
Most AI video looks like AI because people prompt shots, not scenes. The fix that changed my work: - write the script before touching a model - lock one style ref for every shot - cut before motion drifts (~5s) - sound design sells the realism Film grammar > model choice.
11
935
Won the Best Story award in AI category @windscribe WAP Thank you 🖤 Special thanks to my friend @rudolfina13 for encouraging me to participate in this one. My entry 👇
It's time to announce the winners of our $42,000 Human vs AI Animation Contest!! Thank you everyone who participated! It took a long time and many votes to select the finalists, but we did it. And so, without further ado, here are your winners of the WAP Fest!
1
5
403
Three lifetime spots, 48 hours left. I've been in this community since the early days, it's where my work stopped looking like AI and started looking like films. Honoured to be a member. If you're serious about AI animation, go enter. This is the room.
With only 48 hours left, I’m giving away 3 lifetime memberships to the Cartoom Hero course + community ($199 value) 🎁 To enter: • Like this post • Quote repost it • Comment why you want to learn AI animation The 3 best answers win. Good luck. Doors close in 48 hours. 🚨
1
4
321
invideo ran 30+ tests on Google's Omni. The limitation that actually matters got buried in the list: -> Max output is 10 seconds. -> Lip sync drifts after ~6. So what Omni makes is too short to perform on its own. On X, a clip under 10 seconds can miss video scoring entirely, it's in the ranking code. Generation got solved. Stitching clips into something watchable is now the whole game.
Google’s new model Omni is here. It does video, avatars, inpaint, lip sync, and a bunch more, all in one. I spent a day running 30+ tests to figure out what it can do, where it breaks, and whether it's actually production-ready. The unlocks, the ceilings, the specs - full breakdown in the video and the thread.
1
5
884
The fix: treat Omni clips as raw material, not finished posts. Generate 3-4 takes. Cut on the beat right before lip sync drifts (~5s). Stitch to 15-20 seconds minimum so it clears the video-scoring threshold and gives watch time something to measure. The model gives you footage now. The edit is the craft.
196
One prince. One dragon. One kingdom in ash. My entry to @PocketFM_App's #PromptToPixels challenge - a 30-second AI trailer for Brahmand Ka Rakshak, their #9 fantasy show with 55M+ plays and 1,400+ episodes of Hindi mythological epic. The story compresses into one image: a bereft prince meets a tiny dragon in the ashes of his kingdom, and chooses to rise. Honored to have this seen by the jury, including @jdkanani, @oddmanoutnow, and @rupeshkashyap, alongside Ankur Shrivastava, Dr. Ashok Bairagi, Harsh Sharma, and Raj Rishi More Win or not, building this taught me more about AI cinematography than any course could. Grateful to Vaibhav Bachhav for the original story and to Pocket FM for the brief. #AIfilmmaking #PromptToPixels #BrahmandKaRakshak
We just launched PROMPT TO PIXELS - the first AI trailer making challenge. 20-30 sec trailers, $5,600 prize pool, deadline April 20. Your story deserves a screen. Click here: pocketfm.com/prompt-to-pixel… #PocketFM #audioseries #contest
5
532
Same prompt. Two AI tools. The gap is WILD. 👇 Img 1 → NB Pro Img 2 → ChatGPT Image 2 I've been a NBPro loyalist but I can't lie, GPT-Image 2 just cooked. The detail, the hierarchy, the character rendering… it's not even close right now. The AI image race is moving FAST. Are you keeping up? 🧵Prompt in ALT #AIArt #ChatGPT #NanoBananaPro #AITools
🤖 Made with AI
1
7
500
Based on the April 2026 New Yorker investigation by Ronan Farrow & Andrew Marantz - weaponizing the Ilya Memos and the 200-page Amodei notes as source material. Every accusation is a real quote. Sam's denial is too. All sets created in showrunner app and video using SD2 @fablesimulation @dreamina_ai #ExitValley #DreaminaCPP
🤖 Made with AI
5
1
14
354
Test 4 🚨 Complex camera moves. Seedance 2.0 just makes anything look too simple. Single line prompt used : Replace the character in @ Video 1 with the basketball player from @ Image 1. The player’s body proportions, facial features, and athletic build must remain consistent throughout the entire sequence. From court vision to AI POV, following the motion, the rhythm, the flow. Same player. Same movement. New universe. 🏀👁️‍🗨️
Test 2 🚨 Seedance 2.0 can take any video + any reference image and seamlessly swap the person while keeping their outfit, movements, and even changing the language. One prompt. Zero manual editing.
1
1
12
1,553
Tested Seedance 2.0 on a Tollywood-style fight sequence. What surprised me: → Camera actually tracks the action (not random drift) → Impact moments get weight, not just speed → Maintains subject consistency through motion Still not SS Rajamouli, but closer than anything else I've tested.
8
1
16
1,402
Seedance 2 is gone in southeast vpn's (at least for me) The official reason? "Maintenance." The real reason? West discovered it and turned every Marvel scene into anime so fast that China's servers started speaking English and demanding pumpkin spice lattes 🇺🇸☕ They had to pull the plug before the AI started generating Avengers: Endgame 🤣🤣
how to use seedance 2 early: seedance 2 is by far the best video model on the market right now but it doesn't officially release until feb 24th here's how to use it before everyone else it's already available on bytedance's playground go to the link below, create an account, and start using it directly byteplus.com/en/ai-playgroun…
4
5
1,732
i hope they finish all our unfinished queues though (using seedance 2.0 and not the 1.5) 😅
2
473
Test 2 🚨 Seedance 2.0 can take any video + any reference image and seamlessly swap the person while keeping their outfit, movements, and even changing the language. One prompt. Zero manual editing.
Seedance 2.0 lets you feed face reference + 5 outfit images + vid ref and it generated a full fashion video Pose switches. Outfit cuts. Fisheye lens. Ghosting effects. Flickering transitions. All from a single prompt. Precise action replication is HERE. #DreaminaAI #AIlovelanguage #Seedance2 #Seedream5
4
1
12
3,387
Dreamina just quietly dropped the biggest AI video update of 2025. Seedance 2.0 lets you input: → 6 images → 3 videos → 3 audio clips → Text All at once. In one generation. Here's why this changes everything 🧵 #DreaminaAI #AIlovelanguage #Seedance2 #Seedream5
Seedance 2.0 lets you feed face reference + 5 outfit images + vid ref and it generated a full fashion video Pose switches. Outfit cuts. Fisheye lens. Ghosting effects. Flickering transitions. All from a single prompt. Precise action replication is HERE. #DreaminaAI #AIlovelanguage #Seedance2 #Seedream5
2
15
3,923
Character consistency? Solved. One face. Dozens of scenes. Same person every time. Plus: precise action replication from reference videos — camera angles, VFX, pacing, transitions. All replicated exactly. You're not prompting anymore. You're directing.
1
5
477
Seedream 5.0 is even wilder. It searches the web IN REAL TIME before generating images. Ask it to visualize today's news → it pulls live data and generates accurate visuals. Also: smart reasoning that understands physics, logic, and domain knowledge. Fewer AI fails. The combo is lethal: 🟠 Seedance 2.0 = best multi-modal AI video tool 🔵 Seedream 5.0 = first image model with live web search
1
6
627
Seedance 2.0 lets you feed face reference + 5 outfit images + vid ref and it generated a full fashion video Pose switches. Outfit cuts. Fisheye lens. Ghosting effects. Flickering transitions. All from a single prompt. Precise action replication is HERE. #DreaminaAI #AIlovelanguage #Seedance2 #Seedream5
1
1
2
28
10,213