Genova AI

Asked Seedance 2.0 for EXACTLY 10 colored blocks. It panicked and gave me 16. 🫠 I'm stress-testing high-count object permanence. Whenever the subject interacts with one item, the model completely loses its mind trying to track the rest. Has anyone cracked strict entity counts >5 yet? Do I need to force an Image-to-Video workflow, or are there specific parameter tweaks for this? @LudovicCreator @0xInk_ @0xbisc @Framer_X @EHuanglu @johnAGI168 @john87445528 @Dheepanratnam (Prompt in thread 👇)

Grass Comes First made with GPT Image 2 + Seedance 2 #DreaminaCPP @dreamina_ai Prompt below ↓

NO CAGE FOR WUKONG made with GPT Image 2 + Seedance 2 on @raelume Prompt below ↓

Billie Jean Impression made with GPT Image 2 + Seedance 2 on @openart_ai Prompt below ↓

"The Sniper Code" The best way to learn AI animation is by creating 1-minute concept videos. It will level up your skills insanely fast. Here's why👇

No one will believe this is AI. But I Made it with GPT 2 and Seedance 2 on @yapper_so Video Prompt: { "animate": "reference image into a 15s hyper-realistic live football TV broadcast", "visuals": { "shots": [ "wide aerial tracking shot across midfield following play", "side medium shot of intense pressing battle", "attacker dribbles into final third past defender", "curling shot toward top corner, goalkeeper fingertip save", "subtle handheld shake during physical challenge", "crowd erupts with scarves waving", "CUT TO: exact girl from reference image in stadium crowd, oversized club jersey, shocked reaction on jumbotron cam, leaning back slightly, fans blurred behind, floodlights glowing, telephoto lens compression, identity perfectly preserved" ], "consistency": "reference subject perfectly recognizable in final crowd reaction shot", "physics": "realistic ball flight, boot impact, grass divots, net ripple", "grading": "Premier League / Champions League broadcast look", "effects": "anamorphic flares, telephoto compression, natural motion blur" }, "graphics": { "scorebug": "MCI 2-1 REA, clock running from 87:00", "stats_popup": "player name, goals, pass accuracy", "watermark": "top-right sports channel logo", "ticker": "live standings and match updates" }, "audio": { "style": "energetic synced sports commentary with stadium ambience", "dialogue": [ "0-3s: 'Final minutes now! Rodri wide to Foden... cuts inside looking for options!'", "3-7s: 'Madrid press well but City recover — Foden shoots! Incredible save from Courtois!'", "7-10s: 'What reflexes! Look at the crowd reaction — the Etihad is rocking!'" ], "sfx": [ "crowd roar", "boot-to-ball impacts", "net shake", "sliding tackle turf scrape", "distant player shouts" ] }, "specs": { "quality": "photorealistic broadcast realism", "resolution": "1080p 60fps", "style": "UEFA / Sky Sports cinematic broadcast", "lip_sync": "perfect", "artifacts": "none", "identity_preservation": "reference subject likeness must remain exact" } }

画面细节做的挺不错

🎨 KLING 3.0 4K 🎨 Prompt : Third-person dynamic chase perspective, low-angle tracking shot with rapid lateral camera sweeps, a rally race car drifting at high speed along the icy mountain roads of the Monte Carlo Rally, snow-dusted cliffs and tight hairpin turns surrounding the vehicle as chunks of ice spray from its tires, roaring engine revs and gravel crunching beneath the wheels with distant echoes of spectators’ cheers (no subtitles), crisp cold daylight with strong contrasts reflecting off the snow, high-energy cinematic style inspired by premium motorsport commercials. Made in @Kling_ai

A new experiment with Seedance 2.0 on @mitte_ai : FACS. For this videos, I only used FACS codes in the prompt. I didn’t describe the facial expressions in plain language at all. FACS (Facial Action Coding System) is a system for describing facial expressions using individual muscle movements called Action Units (AUs), instead of general emotion labels like “happy” or “sad.” It breaks the face into controllable components such as brow movement, eyelid tension, lip movement and cheek activation. Even though it didn’t follow all 14 Action Units perfectly, it still interpreted most of them surprisingly well. I think this could become especially useful for subtle facial acting, forced smiles, uncanny expressions, mixed emotions and micro-performance details during close-up dialogue shots. You can check the prompt for first 15s, a list of FACS codes generated with GPT Image 2 in the replies.

FLY - Native 4K on @Kling_ai Prompt below 👇

stop prompting AI video generators with just text. it’s basically gambling. if you want actual cinematic control, you have to treat the AI like a human animation studio. exact prompt in the replies 🧵👇 Chat gpt2.0 images /midjourney v 8.1 / seedance 2.0 I wanted a 15-second noir hand-to-hand combat scene. Text-only prompt = messy slop. Visual storyboard + Seedance 2.0 = flawless choreography. @dreamina_ai #DreaminaCpp

Midjourney V 8.1 Cinematic stills Prompts Below 👇

Kitchen Hunt GPT Image 2 + Seedance 2.0 is the best combo right now for speeding up creative workflows. Prompt below ↓

🎨 GPT 2 Serie : Case 10 — Visual Prompt Anatomy Template 🎨 Today : Case 10 : Visual Prompt Anatomy Template A template for showing the structure of a prompt block by block: subject, style, composition, lighting, material, framing, text, constraints, and expected output FInd 4 examples . Prompt template in first comment , other prompts in comments because too long for ALT All visuals are made @pixio_ai 1/6

Prompt Studio: Frozen Titan Awakens with Seedance in @Hailuo_AI Upscaled to 4K with @topazlabs Starlight Precise 2.5 Ultra-realistic arctic wasteland at night, blizzard winds, frozen mountains barely visible through whiteout snow. Scientific expedition placing thermal charges across an ancient glacier. Ice begins cracking in glowing lines beneath their feet. Camera pulls backward fast through snow as an enormous humanoid machine rises from beneath the ice, launching frozen slabs into the air. Helicopters struggle in violent wind overhead. One colossal blue eye ignites through the storm. Final frame: titan fully standing, shadow swallowing the camp.

GPT2 + Seedance 2.0 → Yoga Flow. Controlling pose transition. (Prompt + process in thread) PROCESS: 01. GPT2 Img: Create pose sequence in 02. GPT2 Img: Create base character 03. SD2 Omni-Ref: Animating scene 04. Codex for stitching + frame interpolation While not perfect step-to-step... Still impressive.

Seedance 2.0 - Cinematic Summoning VFX Prompt This prompt generates a multi-cut VFX showcase with a clear progression. It’s designed for realistic 3D with grounded physics, natural character motion and readable, production-style VFX. You don’t need to define the summon, but you can specify what is being summoned if you want control. If you leave it open, the model decides based on the character sheet. You can find gpt image 2 character sheet prompt in the quoted post. Also, this isn’t limited to summoning. You can reuse the same VFX structure for anything, just swap the effect logic. Just change the parts about summoning. Prompt in the replies 👇

@0xbisc @mitte_ai Great work as always 🫡

The Taking of the Relic Made with GPT Image 2 + Seedance 2 on @mitte_ai This time I started with a storyboard, then moved to video. The main reason is cost. Video iterations burn credits really fast. With a storyboard, I can tweak things shot by shot and catch issues early, which is way cheaper. It also makes the whole process more controllable. I can adjust each shot on its own, so the character, composition, and key elements stay consistent. Pacing gets figured out early. Once the storyboard is done, the flow is pretty much set. The video step is just turning those shots into a sequence. Sharing my workflow 👇

Creating videos from such storyboard sketches is simply perfect, isn't it❓🎬 There is no need to write prompts at all, even for video generation.😄

Created this by using a movement sheet as a reference image to animate the dance using Seedance 2.0 + ChatGPT image 2.0 on @yapper_so GPT Image 2.0 Prompt: Dance Sequence Instruction Sheet [VISUAL STYLE] A composition featuring a highly detailed 3D-rendered female dancer. Designed like a professional choreography guide with a technical, diagram-inspired layout. Clean white background, soft studio lighting, and strong contrast to highlight body movement and posture. [GRID LAYOUT] Structured 4×4 panel grid (16 frames total), evenly spaced with thin black divider lines. Each panel is identical in size and clearly numbered from 1 to 16 to show a continuous dance progression. [CHARACTER] Use image1 as the base character. The same female dancer appears consistently across all panels with accurate likeness and proportions. [WARDROBE] The dancer wears a stylish, performance-ready outfit: a well-fitted top paired with a short, flowy skirt. The look should feel modern and visually appealing while still practical for dance movement. Fabric should subtly respond to motion (slight flow and folds), even in grayscale. [PANEL STRUCTURE – EACH FRAME] Top-left: Step number + short dance move title (e.g., “Step 5 – Spin Transition”) Center: Full-body pose capturing a precise moment in the choreography Bottom-left: 3–4 lines of concise instruction describing the move Overlay: Motion arrows and directional guides illustrating how the dancer transitions [MOTION INDICATORS] Incorporate curved arrows for fluid motion, straight arrows for directional steps, and circular indicators for spins or turns. Emphasize rhythm, weight shifts, and body isolation. [RENDER QUALITY] High-detail sculpted 3D style with smooth grayscale shading, subtle shadows, and clean linework. Maintain a polished, concept-art level finish with clarity in every pose. [RESTRICTIONS] No color, no background scenery, no extra characters, no visual clutter, only the dancer and instructional elements.