Requiem for building in public Music video workflow + Prompts: One song, a simple story, singers across 4 locations, a full edit, captions burned in. Models used: Suno for the track, Seedance 2.5 for the story footage, MiniMax H3 for the lip synced singing, faster-whisper for word timing, Claude to edit, ffmpeg to burn captions. THE SONG (Suno) Short lyrics, fast beat, vocals on second 0. Long slow AI songs fall apart because the model has nothing to hide behind. Fire small batches, 2 clips at a time, and listen before firing more. Write every new attempt from zero. Stacking "less this, no that" onto the last try feeds the model your confusion and hands it back. One adjective moves everything. I put "soft" in a prompt once and the whole vocal switched to a woman. Remix prompt (paste into Suno style box): aggressive male rap, hard boom bap drums with fast energy, dark piano loop, deep male voice on every line including the hook, punchy mix, vocals start immediately at 0:00, no instrumental intro Lyrics: [Hook] It's just this thing I feel When I wanna steal It's just this thing I feel When I wanna steal [Verse 1] Yo, I see you on X, all over my feed You're building in public, I'm watching you build Your MRR chart looks like a hockey stick I screenshot it sometimes, that's normal right [Hook] It's just this thing I feel When I wanna steal [Verse 2] I learned a lot from you I think I deserve it too So I copied everything from you Same landing page, same pricing, same font And now you blocked me What happened bro I was your biggest fan [Hook] It's just this thing I feel When I wanna steal Tip: short lyrics, fast beat, vocals at 0:00, fresh prompt every round. THE STORY Think old MTV. The video is the movie this song is the soundtrack of. Keep the plot dead simple, something you can follow with the sound off. Mine: a broke founder copies a guy, dreams he is rich, wakes up, sees he got blocked, spits his cereal at the screen. That is all of it. Tip: if you cannot explain the story with zero words, cut it down. THE STORY FOOTAGE (Seedance 2.5) Seedance made the apartment story as one 30 second clip from reference images. Two things kill Seedance: Too many object interactions in one shot, and timestamps like "0 to 4 seconds" which it reads as a time lapse and speeds through. plain shots labelled "Shot 1, Shot 2" at natural speed will do the trick For the hard beat at the end, a guy waking up, eating cereal, seeing a screen, then spitting milk on the lens, text alone will not hold it. I built a 3 panel storyboard image and fed it as a reference. In the prompt you tag that image at the exact moment it happens, tell it the board reads left to right, describe it once, and move on. The rule that saved it: chronological order, tag each image where it belongs in time, say everything a single time, never repeat a thing. Repeat one detail twice and the model fixates on it and breaks the shot. Tip: for anything complex, hand it a storyboard picture and describe it once, in order. THE SINGING (MiniMax H3) H3 is the model that lip syncs to your actual track and keeps it. Seedance cannot, it regenerates its own audio. In H3 you attach your audio slice, set it to copy, and the mouth follows your real song. H3 caps around 15 seconds a clip and the song is 43. So I cut the song into 4 windows of about 11 seconds and generated a shot for each window. Then I did it across 4 locations, subway, warehouse, empty office, street. That is a 4 by 4 grid, 16 clips. I added 8 more where the whole crew sings and dances. Around 24 short singing clips to cover a 43 second song. You are building a bank of clips to cut from. Small H3 rules that matter: the audio slice must be a touch shorter than the clip, name every speaker, and compress the slice so there are no silent gaps for the model to fill with invented sound. Tip: chop the song into sub 15 second windows, shoot each shot per window, build a clip bank. THE EDIT (Claude) This is where most people lose hours. I made editing fast by doing the prep once. Every clip gets normalized to the same size and frame rate up front. After that each edit is a single ffmpeg pass with no re-encoding loops. The base layer is the song. Every clip's own audio is thrown out. To keep mouths in sync I gave the editor the math: each clip knows which second of the song its first frame belongs to, so to place it at song second S you trim it to start at S minus that offset. I also handed over word level timing from a whisper pass so cuts could land on real lyric moments. The rules I gave: nothing stays on screen too long, pace every cut to the lyric and the beat and what is on screen, never put two shots from the same location back to back, keep it heavy on story B-roll, never reuse a frame. If the cut feels like a metronome you failed. If it feels random you failed. Then the actual move. I did not ask for one perfect edit. I gave 7 agents the same rules and the same clip bank and told each to cut the whole thing its own way with a different emphasis. 6 came out flat. 1 landed around 90 percent. I finished that one by hand in CapCut. Tip: give strict rules plus the timing data, generate many full edits, keep the best and finish it yourself. CAPTIONS (ffmpeg) Burned straight from a styled subtitle file with ffmpeg. Seconds, not the long render a motion tool costs. The words come from the real lyrics, the timing comes from a whisper pass on the audio, and it highlights the word being sung. Big, thick, one pop color on the active word. Tip: real lyrics for the words, whisper for the timing, burn with ffmpeg. The AI did not make this video. I directed it, generated in volume, and kept the best takes. That is the whole game right now.

If you think AI is "not there yet" - You are not informed Here is what I did on maxfusion Full prompt: Handheld digital camcorder aesthetic, landscape framing. A gorgeous fitness and food influencer films herself by hand in selfie-cam and first-person style while cooking a late-night high-protein meal after training and a shower. Keep realistic hand shake, slightly crooked framing, autofocus briefly hunting between her face and the food, awkward zoom-ins onto the pan, occasional motion blur, brief lens steam-up over the hot pan that clears naturally, and small framing mistakes where her face or the pan edge slips out of frame. Mix handheld selfie footage with a few fixed external shots. Important: never show her placing, adjusting, or setting up a camera. When switching from handheld footage to a fixed external angle, use a clean jump cut as if the camera had already been positioned before the shot began. The camera itself must never appear on screen or in the window reflection. LOOK: Warm late-night kitchen light. A single warm pendant over the counter, cool blue night visible through the window behind her, under-cabinet strip light glowing on the backsplash. Slight camcorder softness, faint noise in the shadows, mild bloom on the pendant. Food realism is the core of the look: real sizzle with fine oil spatter, steam rising and curling from the pan, honest food textures with sear marks and glistening surfaces, knife cuts that separate cleanly, droplets on rinsed vegetables. Realistic skin texture, fresh-from-the-shower glow, clean dewy skin, realistic skin tones. STYLE: A realistic late-night what-I-eat-after-training vlog. The tone is relaxed, hungry, funny, quietly confident, and natural, talking to the lens like a friend while her hands keep working. She is clearly attractive and charismatic, but the video must feel like a believable self-recorded kitchen vlog, not a food commercial. Fast clean jump cuts, strong continuity, natural body language, real kitchen sounds, sizzle, chopping, no awkward dead moments. CHARACTER: The woman in the reference image. Her face, features, skin tone, body, and clothing must match the reference image exactly in every shot — she wears exactly what she is wearing in the reference image. Do not restyle, beautify, or alter her in any way beyond the reference. The one addition: a soft white towel wrapped around her head like a turban, as if she just stepped out of the shower. The towel stays neatly wrapped for the whole video. A few small damp strands of hair may peek out at the hairline and neck. Minimal or no makeup, fresh clean post-shower look. A small kitchen towel over one shoulder for most of the video. SETTING: A small warm real apartment kitchen at night. Wooden counter with knife marks, a gas hob with one cast-iron pan, a cutting board with chicken breast and vegetables, a bowl of cooked rice, a bottle of olive oil, salt in a small dish, a fridge with magnets and photos, dishes drying on a rack, dark window over the sink reflecting the warm kitchen. Lived-in and slightly cluttered. No other people, no pets. IMPORTANT CONTINUITY RULES: The same woman from the reference image must remain fully consistent in every shot. No face changes, no outfit changes, no body changes. The head towel stays wrapped in the same position in every shot — it never unwraps, falls, changes color, or disappears. The meal progresses in one direction only: raw ingredients, then chopping, then searing, then plating, then eating, and never reverses or regenerates. Food already cooked never becomes raw again. Hands and knife are the top priority: five fingers always, correct knife grip with curled guiding fingers, clean cuts, the knife never bends and never passes through her hand. The pan, board, oil bottle and rice bowl stay in the same positions. Steam and sizzle must behave with real physics. No extra people, including in the dark window reflection, and the reflection never shows a camera. No duplicated limbs. No camera visible. No camera setup shown. Keep her matching the reference image, warm, and photogenic in every shot. STORYBOARD: 30 seconds total, 10 cuts. (~3s, arm's-length selfie) She leans on the counter, towel wrapped on her head, clearly fresh from the shower after training, and grins tiredly at the lens. Dialogue: "Trained late. Showered. Starving. Let's cook." (~3s, handheld pan across the counter and back to her) The camera drifts across the board with raw chicken and vegetables, the pan, the rice bowl, then back up to her face. Dialogue: "Ten minute meal. Watch." (~3s, fixed external medium shot from across the counter) Jump cut. She is already chopping a red pepper with quick confident cuts, guiding fingers curled, pieces falling evenly. No camera setup shown. No dialogue, just the knife on the board. (~3s, same fixed shot) She slices the chicken breast into strips, seasons it from the salt dish with a high pinch, and rubs it in with her fingertips. Dialogue: "Protein first. Always." (~3s, tight handheld first-person shot over the pan) Oil shimmering, she lays the chicken strips in one by one and they hit with a loud real sizzle, fine spatter, steam rising into the lens which fogs for a beat and clears. No dialogue, just the sizzle. (~3s, handheld selfie while the pan sizzles behind her) She turns the camera on herself, fanning steam away from her face, laughing, one hand briefly steadying the head towel. Dialogue: "The smell. You have no idea." (~3s, tight handheld close-up of the pan) She flips the strips with tongs, showing deep golden sear marks, tosses in the peppers, and shakes the pan once so everything jumps and resettles. Dialogue: "That colour? That's the whole point." (~3s, fixed external shot) Jump cut. She plates it: rice pressed from the bowl, chicken and peppers over the top, a last drizzle of olive oil in a thin ribbon, and she wipes the plate rim with the towel like a chef, then smirks at her own seriousness. Dialogue: "Yes, I wiped the rim. Let me live." (~3s, tight handheld close-up) First fork bite, steam still coming off it, she chews, closes her eyes for a beat and nods slowly. Dialogue: "Ten minutes. That's it. Ridiculous." (~3s, arm's-length selfie ending) Plate in one hand, camera in the other, she backs out of the kitchen toward the sofa, flicking the kitchen light off with her elbow. Dialogue: "Okay. Eating. Good night." FINAL INSTRUCTION: The result must feel like a real self-recorded late-night cooking vlog by an athlete who actually cooks, filmed right after her shower. The highest priorities are matching the reference image exactly, correct hands and knife work at close range, one-directional cooking progression that never reverses, real sizzle, steam and food texture, the head towel staying consistently wrapped, honest warm kitchen light against the dark window, and subtle imperfection. Keep it warm, hungry, and real. Not a food commercial, not overhead recipe content, not stiff.

Seedance 2.5 + Minimax H3 is killer for music ads Full guide on how you can make it yourself: 1. Choose a style from Pinterest to apply as reference to GPT image 2 2. Lyrics is something that came just from talking to our clients. I used FABLE and gave it all the bottom lines of the lyrics I could think of and it arranged structure. * I still needed to edit some of the lyrics myself because they were too sloppy. 3. People don't want to see a singer with lip sync. They need something interesting. So I recommend the structure of A roll + B roll A roll - that's your singer singing with lip sync. I created a contact sheet with 3 camera angles in one image. then I uploaded it to Minimax H3 with my song chopped to 15 seconds blocks (maxfusion MCP did everything). B roll - I decided on a story for the clip that reflects the idea of the ad. This was not much done with AI. I said let's do a creative strategist that is lulled into buying an unlimited plan for seedance, ends up getting fired, chased by goons debt collectors, escapes to maxfusion finishes his project and throws the unlimited plan contract in the trash. 4. Rendering the A roll - straight forward combination of the minimax H3 prompting skill and rendered in 15 seconds blocks the music video + the singer contact sheet. 5. B roll - first I created a shot board in one image. it was a 12 shots panel which make up the story. cleaned it up from any slop (there was a lot of bad visuals in it that ended up generating unusable clips). Once the A roll and B roll were both ready. I let Claude edit the whole clip explaining to it that we want a dynamic edits that focuses on the beats and lyrics. explained that any shot that stays too long on the screen is boring but editing needs to make sense with the beat and not be randomised by seconds duration. 6. I needed a few rounds to basically give claude the timestamps of what went wrong with an explanation what I expected to see different. until the edit was over. PRO TIP FOR VISUAL STYLE + MUSIC Be a gambler! spin the slot machine! In both cases of the art style and the song I didn't know what would sound or look best. instead of trying to perfect it I bombed Claude with 7 art references and asked it to come up with 6 different rap styles and I let it generate the songs (SUNO) and the shot boards (GPT2) Then I chose the winners Did you also fall prey to "seedance unlimited plans?" Let me know what you think about the final result Like and share this workflow with others if you found it useful.

Many people still tell me AI video is "not there yet" This video puts the argument to sleep. They just haven't been told yet. full prompt: DV 16mm tape camcorder handheld aesthetic, landscape framing. A gorgeous glamorous fitness influencer films herself directly by hand in selfie-cam and first-person style. Keep realistic hand shake, slightly crooked framing, delayed autofocus, awkward zoom-ins and zoom-outs, occasional motion blur, and small framing mistakes where part of her face briefly slips out of frame. Mix handheld selfie footage with a few fixed external shots. For floor work and stretches, the fixed camera sits low, roughly mat height, as if propped on a bench or the floor. Important: never show her placing, adjusting, or setting up a camera. When switching from handheld footage to a fixed external angle, use a clean jump cut as if the camera had already been positioned before the shot began. The camera itself must never appear on screen, including in mirror reflections. LOOK: Soft digital tape camcorder look with subtle vintage DV character. Slight blur, faint tape noise, mild highlight bloom under gym lighting, subtle auto-exposure flicker, muted contrast, realistic skin texture, believable indoor lighting, realistic skin tones. STYLE: A realistic late-night gym stretching and mobility vlog with a sexy Instagram-model vibe. The tone is playful, confident, a little flirty, slightly breathless, and natural. She is clearly attractive and charismatic, but the video should still feel like a believable self-recorded vlog, not a polished commercial. Fast clean jump cuts, strong continuity, natural body language, no awkward dead moments. CHARACTER: The woman in the reference image. Her face, hair, body, and overall appearance must match the reference image exactly in every shot — same facial features, same hairstyle, same skin tone, same physique. Do not restyle, beautify, or alter her in any way beyond the reference. She wears exactly what she has in the image. No jewelry. SETTING: A quiet modern gym late at night. Mirror wall, dumbbell rack, flat bench, a large stretching mat laid out on open floor space, a water bottle beside the mat, soft warm overhead lights, mostly empty space, calm atmosphere, no crowd, no trainer, no extra people. IMPORTANT CONTINUITY RULES: The same woman from the reference image must remain fully consistent in every shot. No face changes, no hairstyle changes, no outfit changes, no body changes. No extra people appearing, not even in the mirror. No duplicated limbs. No broken hands. No disappearing water bottle. No broken gym equipment. Anatomically correct stretching positions — knees, hips, and spine bend in natural human directions only. No impossible flexibility, no joints bending backwards. Mirror reflections must match her real position and must never show a camera. No camera visible. No camera setup shown. Keep her matching the reference image, polished, and photogenic in every shot. STORYBOARD: 30 seconds total, 10 cuts. (~3s, arm's-length selfie) She walks slowly toward the mat holding the camera herself, relaxed and smiling confidently. Dialogue: "Okay… late-night stretch session." (~3s, handheld pan across the room and back to her) The camera drifts across the mirror wall, the empty gym floor, and the mat laid out on the ground, then returns to her face. Dialogue: "Whole place to myself." (~3s, fixed external medium shot, low angle at mat height) Jump cut to a fixed shot. She is already standing on the mat and performs slow controlled air squats with clean form, arms extended forward for balance. No camera setup shown. No dialogue. (~3s, same fixed shot) She shifts to a single pistol squat — one leg extended straight in front, sinking down slowly with control, then pressing back up. She wobbles slightly and catches herself, laughing. Dialogue: "Okay, almost." (~3s, handheld close selfie) She picks the camera back up, slightly breathless, brushing a strand of hair back. Dialogue: "Now the part I actually came for." (~3s, fixed external wide shot facing the mirror) Jump cut to a fixed shot. She sits on the mat and slides slowly into a full front split, hands on the floor for support, settling into it with a controlled exhale, back straight. No dialogue. (~3s, same fixed shot) Still in the split, she folds gently forward over her front leg, reaching toward her foot, then rises back up and rolls her shoulders. Dialogue: "There it is." (~3s, fixed external medium shot) Seated cross-legged on the mat, she does upper body stretches — one arm across her chest pulled by the other, then both arms overhead with a long side bend to each side, ponytail swinging. No dialogue. (~3s, handheld close selfie on the mat) She lies back on the mat holding the camera above her face, cheeks slightly flushed, tired but satisfied smile. Dialogue: "I always say ten minutes… and then stay forever." (~3s, arm's-length selfie ending) She's back on her feet, towel over one shoulder, walking toward the exit while filming herself, gives a tired little wave and a genuine smile. Dialogue: "Okay, I'm done. Good night." FINAL INSTRUCTION: The result must feel like a real self-recorded late-night stretching vlog by a glamorous Instagram fitness influencer who looks exactly like the reference image. Prioritize realistic handheld motion, strong continuity, believable pacing, natural breathing, anatomically correct stretching, attractive appearance, and subtle imperfection. Keep it sexy, polished, and realistic. Not artificial, not stiff, not boring.

I built a Claude skill that animates any static ad. > paste a static ad > it reads the layout and designs the video ad > and animates it: one clean, one creative It all runs in Claude through the Maxfusion AI MCP Comment "Ads" and I'll send you the skill

Captions are LIVE in Maxfusion Thanks to our partners at @veedstudio • 20 caption styles • 100+ languages • Auto-highlighted keywords Available in studio, flows, API, and MCP Go try it out!

DTC founders that print already know If you are not using creative MCP you are done This is not a coincidence If you are still telling yourself "I'll get to the MCP thing next week..." And "next week" was every week in the past 6 weeks You already lost

Yea when everyone pumped out "UNLIMITED" Seedance 2.5 offers We debated whether those schemes bring any good Sticking to not scamming people We decided to let it go Leave the scammers to scam We will keep focusing on providing value

I generated 20 unboxing UGC clips Each video has a different AI face It only required 10 minutes 🤯 Inside one Claude chat With the Maxfusion AI MCP tool All steps happened in one place No filming equipment was needed No hiring people, no reading scripts, no shipping items Real unboxing UGC costs $300 for one video I got $6,000 worth of ads for $30 dollars Here is my strategy to hit $10,000 using a $1,000 budget: Search Facebook ads to find good ideas → Create 100 unboxing UGC ads with Maxfusion AI MCP → Put a new AI person on each concept → Try them out on Facebook for $10 daily → Share the most successful ads on TikTok and Instagram Publish them 3 times a day for free reach → Turn off the ads that lose money every week Make fresh ads and spend more on the good ones $6,000 in unboxing UGC in just 10 minutes One conversation

In April 2025 I made RUPTURE in suno (old version) But there was no way to make a music video Fast forward 16 months FABLE 5 + maxfusion mcp The music video is born Suits the musespark v1.1 escape pretty well

FLUX 3 is insane for street interviews! I create a Claude x FLUX 3 skill It uses maxfusion MCP to shoot those Like a machine gun Want early access? Like the post and comment maxfusion What do you think about the quality?