DEV Community: Cittadini Wahler

Why I Started Using AI ASMR Generators for My YouTube Ambient Videos (And What I Learned)

Cittadini Wahler — Tue, 14 Jul 2026 07:34:12 +0000

"A personal story about discovering AI ASMR generators, building an ambient YouTube workflow with them, and what I've learned about creating sleep and relaxation videos the AI way."
It’s been five months since I launched my ambient YouTube channel. Back then, I didn’t own a single camera or mic, and honestly speaking, I was totally flying blind. The only clear goal I had was creating soothing, ASMR-adjacent looped background footage. You know the type—content tons of viewers put on while sleeping, hitting the books, or unwinding with meditation.

What I didn't know at the time was that AI ASMR generators were about to change the entire approach to creating this type of content. And as it turns out, countless other creators are figuring this out right now too. According to Google search trends, searches for "AI ASMR video generator" have been steadily climbing — but there's surprisingly little practical guidance published on how to actually use these tools effectively.

This article is my honest account of why I switched to AI ASMR tools, how the workflow actually works, and what I've learned along the way.

Why I looked into AI ASMR generators in the first place

Making ambient YouTube videos the traditional way — filming or rendering everything manually — is slow. Really slow. Back then, crafting just one 10-minute looping rain window clip would eat up half my whole day. I had to film extra footage, tweak edits, adjust color tones, and layer in matching ambient audio—and all that work was done before I even uploaded it and waited for YouTube’s algorithm to pick it up.

I was publishing maybe two videos a month. Meanwhile, established sleep channels were pushing content daily. I was never going to compete on volume doing it the old way.

So I started exploring AI-driven approaches. Not because I wanted to take shortcuts, but because I saw an opportunity to focus on creative direction — writing better scene ideas, picking the right ambiance, and tweaking details — instead of spending all my time on production grunt work.
[💡 Key insight: Traditional ambient content creation requires video equipment, audio gear, editing software, and hours of post-production. AI ASMR generation collapses this into prompt writing + tool selection. The bottleneck shifts from production time to creative direction.]

The AI ASMR tools I found (and what made one stand out)

I tested several AI ASMR generators before settling into a consistent workflow. Here's the landscape as I found it:

What I eventually settled on was AIforASMR — a dedicated AI ASMR generator that lets me input a text prompt, optionally provide a reference image, and pick from 45+ sound presets like rain, campfire, waves, and forest birds. The output is a complete video with both visuals and ambient audio, ready to upload or embed.

To be clear: I'm not saying this is the only option. But for creating AI ASMR videos for my specific use case — sleep loops, rain ambience, relaxing backgrounds — it solved the workflow problem in a way that general-purpose AI video tools didn't.

What I look for in an AI ASMR video tool

If you're evaluating AI ASMR generators for yourself, here are the criteria I've developed through trial and error:

Sound preset variety. If you want to make rain videos, the tool needs dedicated rain audio — not a generic "nature" sound. Check for specific presets like waves vs. wind vs. forest birds vs. thunder.
Prompt flexibility. Can you describe lighting, camera movement, and scene details? Or are you limited to choosing from dropdowns?
Reference image support. Uploading a mood board image significantly improves output consistency.
Generation speed. For a content pipeline, speed matters. A tool that takes 2 minutes vs. 20 changes your publishing cadence.
Loopability. Sleep videos need to loop seamlessly. Not all generators handle end-to-start transitions well.

My AI ASMR YouTube workflow (step by step)

Here's the exact workflow I use now. It produces a publish-ready AI ASMR video in about 15 minutes per scene:

Step 1: Scene concept

I start with a scene idea. For example: "Rain on a window during golden hour, droplets sliding slowly, warm indoor lighting, soft focus." I write this down as a text prompt.

Step 2: Sound selection

I match the scene with an ASMR sound preset. For the rain-on-window concept above, I select the "Core Sound Rain" preset. For a beach scene, I'd use "Core Sound Waves." The idea is that the visual and audio should feel like they belong together — not like two separate tracks slapped on top of each other.

Step 3: Image guidance (optional)

If I want more control over the composition, I upload a reference photo. This is helpful for product-oriented ASMR scenes or when I need a specific object (candle, skincare bottle, wooden texture) to appear consistently in the video.

Step 4: Generate and review

The AI ASMR generator processes the prompt, image, and sound preset together and returns a video. I preview it, check for motion quality and sound synchronization, then either publish or iterate on the prompt.

Step 5: Upload to YouTube

Final step: add a title with target keywords (e.g., "Rain Sounds for Sleeping — 10 Hours of Cozy Ambience"), write a description, add relevant tags like #ASMR #rainsounds #sleep, and schedule the upload.

What I learned after 5 months of creating AI ASMR videos

This section is perhaps the most useful part — honest takeaways after generating dozens of videos and watching how they performed:

1. Prompt writing is the new editing skill

The single biggest learning curve isn't the tool itself — it's learning to write effective prompts. A prompt like "Cozy rain ASMR" produces something generic. A prompt like "Slow rain on frosted window glass, macro close-up, droplets catching warm golden-hour light, blurry background, soft binaural rain audio, no text, no captions, no watermark, 16:9 landscape, calm movement" produces something you'd actually want to watch.

I eventually started keeping a prompt library organized by ASMR category — rain prompts, wind prompts, candlelight prompts, skincare texture prompts — and that alone has been the highest-leverage productivity investment I've made.

2. The best AI ASMR content is specific, not generic

Videos that describe a specific scene ("Rain on a car windshield at night") consistently outperform generic titles ("Relaxing rain sounds"). Niche ASMR topics — snow walking, keyboard typing, page flipping — have smaller search volume but much higher engagement rates. This is classic long-tail SEO applied to video content, and it works.

3. Sound quality matters as much as visuals

Early on, I made the mistake of focusing all my effort on visual quality and treating the sound as an afterthought. Viewers notice. If the rain sounds like white noise rather than actual rain droplets on glass, people click away.

This is why dedicated AI ASMR generators have a real advantage over general AI video tools — the sound presets are built from actual ASMR audio design, not synthesized as an afterthought.

4. Consistency beats perfection

When I stuck to the old manual production method, I chased flawless footage and only managed to post two videos each month. After switching over to AI-made ASMR clips, I stopped obsessing over absolute perfection and settled for solid, watchable content instead. Now I’m uploading four to five times weekly.
My channel saw way faster growth within just the second month of this new routine than it ever did by month four of my original process. At the end of the day, YouTube’s algorithm heavily favors creators who stick to a steady posting schedule.
[📊 Quick numbers: In 5 months with the AI ASMR generation workflow, I went from 0 to ~2,300 subscribers and my best-performing video (a rain-on-window loop) has ~47K views. Nothing viral — but steady growth that I can attribute to being able to publish consistently.]

Does AI ASMR content actually work on YouTube?

I get this question a lot. The honest answer is: yes, but not the way you might think.
It doesn't work if you:

Generate generic, low-effort scenes without thoughtful prompts
Ignore audio quality (sound is non-negotiable for ASMR and sleep content)
Expect overnight viral growth (sleep channels are a slow-burn game)

It does work if you:

Invest time in prompt writing and scene design
Match the right sound preset to each visual scene
Publish consistently and treat it like a real content pipeline
Optimize titles and descriptions for search (this applies to AI-generated content too)

One thing I'll say definitively: the line between "authentic" and "AI-generated" matters less to viewers than most creators think. What matters is whether the video feels calming, whether the sound is immersive, and whether the loop works. If you hit those three things, viewers don't care what tools you used to create it.

What's next for my AI ASMR workflow

I'm five months into this experiment, and I'm still exploring. Some things I'm working on:

Building a broader prompt library organized by ASMR category (tactile, nature, mouth sounds, product textures, etc.)
Experimenting with thumbnail design — AI-generated video frames make great thumbnails with minimal editing
Trying AI ASMR video generation for shorter formats: YouTube Shorts and Instagram Reels

If you're curious about AI ASMR generators, I'd recommend skipping the general-purpose tools and looking for ones built specifically for this niche. You'll save yourself weeks of trial and error.

If you have questions about the workflow or want to share your own experience, drop a comment below. I'm still learning, and conversations with other creators have been the most valuable part of this process.

I Tried AI ASMR Video Generators for 30 Days — Prompt Examples, Tools, and YouTube Results

Cittadini Wahler — Wed, 08 Jul 2026 07:22:28 +0000

My apartment faces a busy street, so forget about recording anything quiet.
But I wanted to make ASMR videos anyway. So I spent 30 days testing AI ASMR video generators to do the heavy lifting.
Four tools, 80+ prompts, ~200 clips, one YouTube channel, $34.90 total. Some of it worked. Some of it was hilariously bad. Here is the full story.

Why ASMR and AI are a surprisingly good match

Look, I am not going to pretend I picked ASMR for artistic reasons. I picked it because ASMR content is weirdly well-suited to what current AI generators can actually do.
Think about it. Most ASMR scenes are slow or static: rain sliding down a window, a candle flame doing its thing, a skincare close-up. No explosions, no action sequences, no characters running around. AI video models are terrible at complex motion, but they can handle "rain on glass" surprisingly well.
And here is the thing about ASMR viewers — they care about mood first. Soft focus? That is a vibe. Warm lighting with a bit of blur? That is atmospheric. The things that make AI video look "off" in a normal context actually work in ASMR.
Plus the production bar is ridiculously low. A 10-hour sleep video can be a single 30-second clip looped with some rain audio on top. No script, no plot, no voiceover.
YouTube is full of "rain sleep 10 hours" and "candle ambience" videos pulling consistent search traffic. You do not need subscribers for discovery.

The tools I used (and the ones I wasted time on)

I tested four AI ASMR video generators. Let me save you the trouble.
Runway Gen-3 — Beautiful quality, genuinely impressive. But clips capped at 10 seconds and zero ASMR-specific templates. You are on your own for prompt engineering.
Pika Labs — Free and fast. That is the good news. The bad news: watermarked clips and 8-second maximum length. Only useful for testing ideas.
Kling AI — Surprisingly good at textures but completely inconsistent with slow motion. Half my candle clips looked like the flame was in a wind tunnel.
AIforASMR.com — This one surprised me. Among all the AI ASMR video generators I tested, this was the only one with 45+ built-in ASMR scene templates and clips up to 30 seconds. The Starter plan is $19.90/month for 1,600 credits. The template library was the killer feature. After week one, I stopped using everything else.

30 days, week by week

Week 1 — Peak overconfidence

My very first prompt with these AI ASMR video generators: "Rain video ASMR." What I got back was 6 seconds of illegible gray static. Looked like TV snow from the 90s.
Turns out, a good prompt needs four things: a subject, the camera position, lighting and tone, and motion quality. The first prompt that actually worked:
"Static macro shot of rain droplets sliding down dark bedroom window glass, soft city lights blurred in the background, slow and calm motion, warm amber indoor light reflection visible on the glass surface"
Still not perfect — water physics looked fake — but the mood was right.
Scene-by-scene report card: Rain on glass and skincare textures worked well across all tools. Candle flames needed specific lighting. Fabric was inconsistent. Water ripples were a disaster. Mouth movements? Do not bother.

Week 2 — Templates saved my sanity

60+ generations → 12 usable clips (20% yield). Writing every prompt from scratch was exhausting.
Before templates: 45 minutes of tweaking per usable clip. With templates: ~30 seconds. Click a preset, adjust 2-3 words, generate. Yield jumped to ~70%.
Week 3 — Wait, clips are only 10 seconds?
All the AI ASMR video generators I tested capped clips at 5-30 seconds. YouTube sleep videos run 2-10 hours. Fundamental mismatch.
FFmpeg merge worked but every clip had different brightness. Crossfade transitions in DaVinci Resolve (free) smoothed it out. Best fix: spend more credits for 30-second clips and only need 4-6 per loop.
One hard lesson: ASMR without good audio feels hollow. I matched rain sounds to rain visuals, fire crackling to candle scenes. Royalty-free tracks from Pixabay and YouTube Audio Library.

Week 4 — YouTube results

The rain video got 63% of all views. Marked as AI-generated, still appeared in search within 48 hours. No suppression.
Total cost: $34.90 (AIforASMR Starter plan $19.90 + $15 top-up). Everything else was free.

What I would do differently

Start with templates. Week 1 was basically a waste. 3x more content with the same credit budget.
Pick audio first. Matching audio to clips is harder than generating visuals to match pre-selected tracks.
Batch-generate. 15-20 sessions back to back > 3-5 per day for consistency.
Skip free tiers. Watermarked 5-8 second clips are not publishable.
Use image-to-video for consistent look. Templates give variety; reference images give consistency.

Prompts that actually work (copy these)

These worked across multiple AI ASMR video generators. Swap the bracketed parts for your scene.
Rain / Water
"Static macro shot of [rain droplets / condensation beads] sliding down [dark bedroom / car] window glass, [city lights / street lamps] blurred in bokeh background, slow gentle motion, [cool blue / warm amber] interior reflection"

Candle / Fire
"Single [white / pillar / jar] candle flame on [wooden / marble] table, complete darkness surrounding, micro-motion flame sway, warm amber glow, static camera, 24fps"

Skincare / Texture
"Macro close-up of [serum / moisturizer / oil] spreading on [fair / deep] skin, soft window lighting, glossy reflective texture, slow circular motion, dewy finish"

Should you try this?

Honest answer: yes, if you want to test the format without buying gear.
$35 got me 4,200+ views on a rain loop. At $2-5 CPM, that is about $8-20/month from one video. Build up 30-40 videos and it could turn into a modest side income.
But AI handles production, not audience building. You still need good SEO, regular uploads, and patience.
Grab a rain template from AIforASMR.com, generate 5-10 clips, add royalty-free audio, and upload. First video in about an hour.

TL;DR
30 days, $34.90, no equipment. 4,200+ views. Rain > candle > skincare > fabric. Templates beat raw prompts by a massive margin. Clip length is the main constraint, fixable with crossfades or higher credits.

50 ASMR Video Prompts for AI Generators (Copy & Paste Ready)

Cittadini Wahler — Tue, 07 Jul 2026 07:42:15 +0000

Writing an ASMR video prompt sounds easy. "Calm rain on window" — four words, what could go wrong? Then the AI generates a grey smear. The difference between unusable noise and something you could publish comes down to three things: lighting description, negative constraints, and camera framing.

There are fifty prompts here, split into ten ASMR categories. They work great on AIforASMR, Runway, Pika, Luma, and any other AI video generators. Every prompt notes what to exclude: no visible text, no watermarks, smooth endless loops.

Structure: [Subject] + [Motion] + [Texture] + [Lighting] + [Camera] + [Negative constraints]

Ever wonder why five separate details make such a difference? Write “flickering candle” but forget to mention its lit wick, and the whole shot feels flat. Just typing “rain hitting a window” without specifying framing could give you a full house wide shot instead of that tiny close-up you pictured. Covering every key detail leaves zero room for the AI to misread what you’re trying to create.

🌧️ Rain & Water (6)

Most reliable category. Water textures, reflections, and droplet physics render best. The templates at AIforASMR include pre-configured rain scenes with matching sound presets.
1. Rain droplets sliding down a window pane in soft grey morning light, macro close-up, gentle bokeh background, no text, no watermark, seamless loop

2. Heavy rain falling on a dark green leaf, droplets beading and sliding off the tip, deep green background, macro focus, natural daylight, no text, no watermark, seamless loop

3. Slow-motion water droplets falling into a small puddle, concentric ripples spreading outward, overcast sky reflection, macro from above, crisp focus on ripple center, no text, no watermark, seamless loop

4. Gentle stream flowing over smooth river stones in a forest, clear water, moss-covered rocks, dappled sunlight through canopy, slow steady movement, no text, no watermark, seamless loop

5. Rainwater dripping from a metal gutter edge, each drop catching dull grey light, macro close-up, slow drip rhythm, concrete background blurred, no text, no watermark, seamless loop

6. Morning dew on a spider web, tiny water beads suspended on silk threads, soft golden sunrise light behind, macro extreme close-up, gentle breeze sway, no text, no watermark, seamless loop

🌊 Ocean & Waves (5)

7. Gentle ocean waves rolling onto a sandy shore, white foam receding slowly, warm golden hour light, wide horizontal view, slow natural rhythm, no text, no watermark, seamless loop

8. Close-up of water lapping against weathered wooden pier pillars, small waves breaking, soft reflections, blue-green water, afternoon light, no text, no watermark, seamless loop

9. Ocean surface seen from above, slow rolling swells, sunlight glittering on water in moving patterns, deep blue and silver tones, gentle camera drift, no text, no watermark, seamless loop

10. Tide pools at sunset, small waves washing over rocks covered in sea anemones, warm orange and pink sky reflection, macro focus on water surface, no text, no watermark, seamless loop

11. Mist rising off the ocean at dawn, calm sea surface, soft grey-blue tones, distant horizon barely visible, slow horizontal pan, no text, no watermark, seamless loop

🔥 Candlelight & Fire (5)

Best for sleep and meditation. Warm-toned scenes are forgiving — even imperfect generations look good in amber light.
12. A single white candle burning in complete darkness, flame flickering gently, warm orange glow against black background, macro close-up, slow natural flicker motion, no text, no watermark, seamless loop

13. Multiple candles arranged on a wooden table, flames casting warm amber shadows, soft golden backlight, shallow depth of field, slow subtle movement, no text, no watermark, seamless loop

14. A campfire in a stone fire pit, flames rising and falling, embers floating upward into dark night sky, warm orange and red tones, steady medium shot, no text, no watermark, seamless loop

15. Hot wax dripping down the side of a candle, pooling at the base on a ceramic holder, warm amber glow, macro close-up, slow motion, no text, no watermark, seamless loop

16. A match being struck and igniting, quick bright flame, slow burn down the matchstick, dark background, macro close-up, warm orange light, no text, no watermark

🌲 Nature & Ambient (6)

Strong for long-form YouTube backgrounds. Green tones + slow drift = natural loops viewers won't notice repeating.
17. Sunlight filtering through dense forest canopy, rays visible through mist, deep green and gold tones, slow upward camera drift, no text, no watermark, seamless loop

18. Tall grass moving in a light breeze, golden hour light on each blade, shallow depth of field, macro close-up, slow swaying motion, no text, no watermark, seamless loop
_
_19. Snow falling gently in a pine forest, large fluffy flakes drifting down, soft blue-white light, evergreen trees barely visible, slow peaceful motion, no text, no watermark, seamless loop

20. Autumn leaves floating on a calm pond, slow circular movement, warm red and orange colors reflected on water surface, afternoon light, overhead view, no text, no watermark, seamless loop

21. Cherry blossom petals falling slowly through the air, pale pink against soft blue sky, gentle breeze carrying petals sideways, shallow depth of field, slow motion, no text, no watermark, seamless loop

22. Bamboo forest in light wind, tall green stalks swaying gently, dappled light through dense leaves, slow upward pan, natural green tones, no text, no watermark, seamless loop

🪵 Wood & Tactile (5)

Second most popular ASMR category on TikTok and Shorts. Camera must be close enough to show texture.
23. Wooden chopsticks tapping slowly on a polished bamboo bowl, clean crisp motion implied, warm amber lighting, macro close-up, shallow depth of field, no text, no watermark, seamless loop

24. Fingers running slowly over textured wooden surface, close-up of grain lines and knots, natural wood tones, soft window light, macro focus on fingertips, no text, no watermark, seamless loop

25. Wooden blocks being stacked and gently placed on a table, each piece catching warm afternoon light, shallow depth of field, slow deliberate motion, no text, no watermark, seamless loop

26. A wooden comb being slowly drawn through long hair, visible static and gentle movement, warm backlight highlighting hair texture, macro on comb teeth, no text, no watermark, seamless loop

27. Carving a piece of wood with a small knife, fine wood shavings curling off, close-up of hands and tool, warm workshop lighting, slow careful motion, no text, no watermark

🍔 Food & Drink (5)

High engagement on Instagram Reels. Macro texture + implied taste makes these the most shared ASMR clips.
28. Ice cubes slowly melting in a crystal glass filled with water, condensation forming on the glass surface, cool blue lighting, macro close-up, slow melting motion, no text, no watermark, seamless loop

29. Hot coffee being poured into a ceramic cup, steam rising in visible wisps, dark brown liquid swirling, warm morning light, close-up from above, slow pour motion, no text, no watermark, seamless loop

30. A knife slicing through a crisp apple, blade cutting cleanly through the flesh, juice visible on the cut surface, bright natural light, macro on blade contact point, no text, no watermark

31. Honey slowly drizzling from a wooden dipper onto golden pancakes, thick viscous stream, warm morning lighting, macro close-up, slow continuous pour, no text, no watermark, seamless loop

32. A carbonated drink being poured over ice, bubbles rising rapidly, condensation on the glass, cool tones with backlighting, macro close-up, fizzy motion, no text, no watermark, seamless loop

✋ Touch & Hand Sounds (4)

33. Fingers tapping slowly on a glass table surface, one finger at a time in sequence, sharp clean surface, macro close-up from above, natural daylight, no text, no watermark, seamless loop

34. Pages of a book being turned slowly, paper texture visible in macro, warm reading lamp light, fingers gently lifting each page, shallow depth of field, no text, no watermark, seamless loop

35. Fabric folding slowly on a smooth surface, soft cotton texture in macro, natural light showing fabric weave, slow crease and unfold motion, no text, no watermark, seamless loop

36. Fingers gently tapping on a wooden desk surface, alternating fingers in a slow rhythm, warm desk lamp light, macro close-up side angle, no text, no watermark, seamless loop

📄 Paper & Writing (4)

37. A fountain pen writing slowly on textured paper, ink flowing onto the page, elegant cursive script forming, warm desk lamp light, macro on nib and paper contact, no text, no watermark, seamless loop

38. Scissors cutting through a sheet of paper, clean straight cut line, paper texture visible in macro, bright daylight, moderate speed cut, no text, no watermark

39. Pages of a notebook being flipped through rapidly, paper sounds implied, natural light, medium close-up, fast page-turn motion, no text, no watermark

40. A pencil drawing on rough paper, graphite leaving a trail, shading strokes building up texture, warm light from the side, macro on pencil tip, no text, no watermark, seamless loop

🎵 Instruments (4)

Niche but high-retention — viewers watch to see the next strike or pluck.
41. A crystal singing bowl being struck with a wooden mallet, visible vibration on the bowl surface, soft ambient light, macro slow-motion, golden reflections on the bowl, no text, no watermark, seamless loop

42. Kalimba being played slowly, metal tines being plucked one by one, visible vibration of each tine, warm natural light, macro close-up on fingers and tines, no text, no watermark, seamless loop

43. A violin bow moving slowly across strings, horsehair contacting the string, visible fine vibration, warm concert hall light, macro extreme close-up on contact point, no text, no watermark, seamless loop

44. Raindrops falling on a metal wind chime, each drop causing a gentle movement, grey rainy light, outdoor vertical composition, macro close-up on chime tubes, no text, no watermark, seamless loop

🔇 Daily Objects (6)

Most varied category. Everyday objects produce the most relatable ASMR viewers recognize the sound and motion.
45. A typewriter key being pressed slowly, the arm swinging up to strike the paper, mechanical linkage visible, warm library light, macro on key mechanism, no text, no watermark, seamless loop

46. A zipper being pulled slowly up and down, teeth interlocking and separating, denim fabric around the zipper, natural daylight, macro close-up, slow even motion, no text, no watermark

47. Old-fashioned clock pendulum swinging back and forth, brass surface catching light, dark wooden background, steady rhythmic motion, warm amber light, no text, no watermark, seamless loop

48. A matchbox being slid open and closed slowly, cardboard texture, fingers on the box, warm indoor light, macro close-up, slow sliding motion, no text, no watermark, seamless loop

49. A ceramic mug being gently rotated on a wooden table, hands cupping the mug, steam rising, warm morning light, medium close-up, slow rotation, no text, no watermark, seamless loop

50. A paintbrush being dipped into water, pigment swirling off the bristles, visible color mixing in the water, natural light from above, macro on brush and water contact, no text, no watermark, seamless loop_

Quick Reference

How I Use These (You Can Too)

Pick a starter category from the table. New channel? Rain or Candlelight — they fail the least. Or just use the pre-built templates at AIforASMR and skip the setup.
Drop in the full prompt. Don't trim the negative constraints — that's the difference between a clean loop and random branding.
Fix the lighting (see table below). One variable, biggest impact.
Set ratio: 16:9 (YouTube), 9:16 (Shorts/Reels), 1:1 (Instagram).
Budget for 3–5 rounds per scene. Change one thing at a time.

Lighting Tweak Reference

Three Prompt Hacks I Wish I Knew Earlier

Front-load the most important element
AI models pay more attention to tokens at the beginning of the prompt. If texture is the star — wood grain, water droplets — start with it, not the lighting.
Match sound to scene
AIforASMR includes matching sound presets for every category — rain sounds for water scenes, tapping for wood. If your tool doesn't have this, layer ambient audio in post.
Save your winning prompts
Keep a document with one line per prompt — the text that worked, the settings, and the lighting tweak. You'll reuse them every time you start a new scene.

So that's it. 50 prompts. 10 categories. No more grey potatoes.
Rain sliding down glass at golden hour. Candle flicker in the dark. Fountain pen tracing cursive across paper. You have everything you need to make these.
Pick one category tonight. Gen three scenes. Post the best one tomorrow. In two weeks you'll have more publishable loops than you know what to do with.