Catcher.

web3 Explorer👨‍💻

[Catcher. @0xCatcher_]

Small creators: Finally found a platform 🥹
@GimiApp: 😬

[Fikkkfx @FikkkFX]

「冥府の門は、貴様ごと断つ！」

Dark Fantasy 2D Sakuga Key Visual.
Obsidian Torii Chasm under the Crimson Eclipse.

Visual generated with Master Profile 02 (Midjourney V6.1).

Made with Seedance 2.5 on Flova

@Flovaai @Flovaai_Japan
#Flovaai #FlovaCPP

[razornero @Danzeronero]

Made this drifting car with seedance 2.5 1080p on @fal

Enjoy drift. 😉

[Fikkkfx @FikkkFX]

Always trying to create new styles

Made with Midjourney v 8.2 on @atlas_cloud_ai

[razornero @Danzeronero]

Made with Seedance 2.5 in 1080p using @Flovaai Agent , an original open-world escape mission concept , cinematic city pursuit, player-view action, fictional mini-map HUD, wanted stars, and a full-throttle getaway.

#flovacpp #flovaAi

[Fikkkfx @FikkkFX]

Weekend mode: explore new style

Testing out fresh styles with MiniMax Design, sometimes you just gotta switch things up.

@Hailuo_AI
#MiniMaxDesign

[razornero @Danzeronero]

One brief ! Thousand ideas , a hundred submission.
@GimiApp

[Fikkkfx @FikkkFX]

Just test style from @imhealmachine

Made with Midjourney v 8.2 on @TopviewAIhq @TopviewAIJP

Thanks to @imhealmachine, I really like this style😍

[razornero @Danzeronero]

Been experimenting with music-driven motion in MiniMax H3, and this one turned out pretty clean.

I wanted to see how far I could push still images with sliced panels, kinetic typography, glitch transitions, and rhythmic motion.

15 seconds.
A few images.
One beat.

Made with MiniMax H3 on @atlas_cloud_ai

[Fikkkfx @FikkkFX]

MiniMax H3 is kinda wild ngl, one prompt and it turns into a whole cinematic trailer with kinetic pop art energy.

AI video, but make it feel like a movie drop.

Made with MiniMax Design @Hailuo_AI
#MiniMaxDesign

[Fikkkfx @FikkkFX]

Tried @zeely_ai for making UGC and ngl, I wasn't expecting it to be this easy.

You just drop your product link, pick an avatar (or make your own if you want), and add a reference too if that's your thing. Then generate the image, generate the video, and that's literally it. One click.

All the usual hassle of finding talent, location, shooting? Gone. You just need a laptop and a few minutes.

[razornero @Danzeronero]

I just made a short movie with Seedance 2.5 using @fal Agent, and honestly… this feels different.

I started from the character sheet, built the characters, developed the scenes, and turned everything into a short movie.

I didn’t need to write complicated prompts or explain everything like a technical brief.

I could just talk to the Agent like I’m talking to a friend:

“Okay, let’s make this character look like this.”

“Now create the next scene.”

“Change this part, keep the character consistent.”

And it just understood what I wanted.

That’s what impressed me most.

AI filmmaking is slowly moving from “learning how to prompt” to simply knowing what you want to create and having an Agent help you make it happen.

This short movie was made with Seedance 2.5 + @fal agent .

[Fikkkfx @FikkkFX]

20 MOTIVATED CUTS IN 30 SECONDS?! Pure 2D Anime Taijutsu Overdrive!

One single opening is all it takes to turn the entire fight around. Put your headphones on for the full impact!

Generated with @renoiseai

#Seedance

[Fikkkfx @FikkkFX]

Can you fit 30 camera cuts into a single seamless 30s anime sequence? Challenge accepted.

Featuring elite operative KAI executing a relentless 5-movement Gun-Fu barrage, from 14m/s knee-slides to bullet-time solar cascades.

Directed with full 2D Sakuga cel-shading & motivated focal transitions.

@TopviewAIhq @TopviewAIJP
#Topview30sChallenge #TopviewAI

[Apil @apilpirman]

Hari ke-3 ngetest agent skill bareng @fal.

Jadi, FAL itu agent yang udah punya orkestrator, context memory, dll.

Cuma, buat bikin video dan gambar, kita tetap butuh file skills dalam format ".md".

Nah, yang di bawah ini adalah workflow yang gw bikin sendiri yg berasal dri file skills itu.

[razornero @Danzeronero]

Same prompt. Two different models.

Seedance 2.0 vs Seedance 2.5

The upgrade in detail, lighting, and motion is insane.
2.5 feels way more cinematic .

Try this compare model on @atlas_cloud_ai .

Which one do you prefer?
#Seedance25 #Seedance20

[razornero @Danzeronero]

Built this car test-drive sequence with Seedance 2.5 using a video reference in @atlas_cloud_ai .

From POV startup → racing circuit → high-speed drift → drone slow-mo → back to the garage.

Still exploring how far realistic car cinematics can go with AI.

First prompt : Create a 30-second ultra-realistic POV car reveal using the provided car image as the exact reference.

The camera is the character's eyes, filmed like real handheld footage inside a real garage. No face visible.

0-6s: Walk toward the parked car and say:
"Alright... this is my new ride."
Move around the front, showing realistic paint, anime livery, headlights and body details.

6-12s: Open the hood and reveal the realistic performance engine. Camera moves closer while saying:
"Let's see what's hiding under the hood."

12-18s: Walk to the driver's side, open the door and look inside. Show the steering wheel, dashboard, racing seats, gear shifter and interior. Say:
"And the inside... this is where it gets serious."

18-24s: Walk toward the rear and show the spoiler, taillights, diffuser and exhaust. Say:
"And just look at the back..."

24-30s: Walk backward and reveal the entire car. The hood and driver's door remain OPEN, showing both the engine and interior. Say:
"Yeah... this is exactly how I imagined it."

STYLE:
Photorealistic live-action automotive footage, realistic handheld POV, natural walking movement, realistic lighting, reflections, shadows, autofocus and depth of field.

Keep the exact same car, livery, wheels, body kit and proportions throughout. No CGI, no 3D render, no cartoon, no morphing, no floating camera.

Duration: 30 seconds
Aspect ratio: 16:9

Second prompt :

Create a 30-second ultra-realistic automotive test-drive sequence using the provided car image as the exact vehicle reference.

The video continues directly from the previous scene. The same car, same garage, same body kit, same white-and-red anime livery, same wheels, same spoiler and same proportions must remain perfectly consistent.

0-6s: POV enters the car inside the garage. Hands grip the steering wheel, press the start button, and start the engine. Say:
"Alright... let's see what she can do."

6-11s: Shift into gear and drive out of the garage. Realistic interior POV, engine sound, steering movement and road vibration.

11-17s: Cut to a realistic exterior tracking shot. The car enters a closed racing circuit and accelerates hard.

17-23s: Drone camera follows from above as the car performs an aggressive controlled drift around a corner. Tire smoke, counter-steering and realistic weight transfer. Brief cinematic slow motion at the peak of the drift.

23-27s: Low rear tracking shot as the car exits the drift and accelerates hard. Brief subtle blue exhaust flames.

27-30s: POV returns to the garage. The car drives back inside and stops. Say:
"Yeah... she's ready."

STYLE:
Photorealistic live-action automotive footage, realistic car physics, natural lighting, realistic tire smoke, authentic engine sound, cinematic camera movement, realistic drone footage.

Keep the exact same car, livery, wheels, body kit and proportions throughout.

No CGI, no 3D render, no cartoon, no morphing, no unrealistic physics.

Duration: 30 seconds
, 16:9.

If you lazy to read the prompt just DM me i'll give it to you

[Fikkkfx @FikkkFX]

Testing out pure 2D Sakuga Anime Action using Seedance 2.5 at 1080p!

The dynamic camera tracking, elemental particle trails, and impact frames turned out incredible.

Made with Flova @flovaai @Flovaai_Japan
#Flovaai #Flovacpp

[Fikkkfx @FikkkFX]

The line is open!

@GimiApp

[Fikkkfx @FikkkFX]

Tested making a shoe ad in full kinetic pop art typography, 15 seconds, multi-clip on MiniMax Design.

Turns out it can chain clips together smoothly without manual editing. Comic-halftone bursts + exploding typography, wild result for one generation.

@Hailuo_AI #MiniMaxDesign

[razornero @Danzeronero]

One brief , a thousand idea .

Hundred of submisson in 24 hours , only pay for what you want.

@GimiApp