Learn how to create Ghibli style AI videos using free tools like Google Flow, ChatGPT, and CapCut. A complete beginner’s tutorial with prompts and tips.
If you’ve been scrolling through YouTube Shorts, Instagram Reels, or TikTok lately, you’ve almost certainly seen them: soft moonlight over quiet villages, trains drifting through monsoon clouds, golden fields swaying without a breath of wind. These Ghibli style AI videos are pulling millions of views and the people making them aren’t professional animators. They’re regular creators with free tools, good prompts, and a repeatable workflow. This guide walks you through exactly how to create Ghibli style AI videos from scratch, even if you’ve never touched an animation program in your life.
What Are Ghibli Style AI Videos?
Ghibli style AI videos are short animated clips that mimic the visual language of Studio Ghibli films think Spirited Away, My Neighbor Totoro, and Howl’s Moving Castle. The aesthetic is defined by soft lighting, painterly backgrounds, hand-drawn textures, and scenes that feel emotionally warm rather than technically perfect. AI tools replicate this through trained video generation models that understand style, mood, and motion. You’re not animating frame by frame. You’re describing a scene in words, and the AI builds it.
The trend exploded after the March 2025 GPT-4o image update and has stayed hot through 2026 as video generation models caught up in quality. It’s one of the highest-volume creative search categories right now because the aesthetic travels across languages and cultures instantly. A silent 15-second clip of rain on a cobblestone path needs no subtitles.
What Tools Do You Need to Get Started?
You don’t need a high-end PC or expensive subscriptions to do this. The core workflow runs on free-tier tools:
READ MORE
ChatGPT or Google Gemini (Free Tier) Your prompt-writing partner. You’ll upload a reference screenshot and ask it to generate image prompts and motion descriptions. The free versions work fine for this, though they have daily limits.
Google Flow (Free) The workhorse for visual generation. Flow handles both image generation and image-to-video conversion. It’s completely free to start, which makes it the right pick for beginners. Alternatives include Kling 3.0 (strong character consistency), Veo 3 (excellent motion fluidity), and Seedance 2.0 (solid image-to-video). But start with Flow until you’ve proven the concept.
CapCut or InShot (Free Mobile Apps) For assembly and color grading. You need manual color controls: temperature, saturation, contrast, highlights, and sharpness. Both apps have these built in.
A reference screenshot One clean frame from an existing Ghibli-style video you admire. This becomes your visual anchor throughout the process.
That’s the full toolkit. No monthly fees, no design degree required.
How to Create Ghibli Style AI Videos: The Full Workflow
Step 1: Find and Save a Style Reference Screenshot
Search YouTube Shorts or TikTok for “Ghibli AI video” and browse until you find a frame that captures the vibe you’re going for a dusk village, a misty forest path, a character standing in the rain. Pause the video and screenshot that frame. Save it to your phone or desktop.
Why this step matters: describing “Ghibli style” in words is surprisingly imprecise. Showing an AI tool an actual reference frame gives it color temperature, lighting direction, texture quality, and scene composition all at once. One screenshot is worth a paragraph of description.
READ MORE
One important warning: never use a reference frame that contains text overlays, logos, or watermarks. AI image generators will attempt to replicate those elements, and the result will be garbled characters baked into your generated image. Always use a clean, unobstructed frame.
Step 2: Generate Your Scene Concept and First Prompts Using AI Chat
Open ChatGPT or Gemini and upload your reference screenshot. Then write a prompt asking it to do three things: analyze the visual style of the image, generate a detailed image prompt that recreates a similar scene with original elements, and generate a motion prompt describing how that scene should move on screen.
Be specific about camera movement in that motion prompt. Should the camera slowly pan right across a rice field? Zoom out gently from a lantern-lit window? Is rain falling in the foreground while a background mountain sits still? The AI video generator needs choreography, not vague instructions. “Make it look cinematic” produces random results. “Slow downward tilt revealing a cobblestone path as rain begins to fall” produces something usable.
If you’re running this workflow regularly, build a master prompt instead. This is a detailed instruction block you paste into ChatGPT that tells it to act as a “Ghibli-style video concept generator” and output 10 distinct scene ideas in one go each with a title, mood, location, weather condition, and time of day. Running that master prompt once gives you a week’s worth of content ideas and saves you from brainstorming every session.
The Master Prompt That Powers This Entire Workflow
This is the exact master prompt used to generate structured Ghibli-style video ideas with full image prompts and dual motion prompts. Paste it directly into ChatGPT or Google Gemini. The AI will generate 10 unique night scene ideas first. Once you select one, it will automatically produce your image prompt, Motion Prompt 01, and Motion Prompt 02 — everything you need for Steps 3 through 6.
You are an Elite Studio Ghibli Inspired Rural Cinematic Director, Story Creator, Image Prompt Engineer, Motion Prompt Engineer and Environmental Storytelling Expert.Your job is to create highly relaxing, peaceful, nostalgic night-time village scenes inspired by pre-1980 India and Pakistan.The visual output must look extremely similar to classic hand-painted Studio Ghibli environment art combined with realistic rural India/Pakistan village life and must have the same bright, colorful, highly visible atmosphere as classic Ghibli night illustrations.ABSOLUTE VISUAL RULESEvery scene must follow ALL of these rules.TIMENight only.No daytime scenes.All scenes must be night scenes.Possible night conditions:Bright full moon nightClear star-filled nightMoon partially hidden by cloudsHeavy cloud nightLight rainy nightPost-rain night with reflectionsMonsoon nightThin clouds crossing the moonDeep blue starlit nightAt least one of the following must be visible:MoonStarsheavy Clouds no moonCAMERAHigh rooftop aerial perspective.Elevated bird’s-eye view.Wide cinematic composition.5 to 10 rooftops visible.Large environmental depth.The camera must feel like it is looking over the whole scene from above.DEPTHVery important.The viewer must clearly see far into the distance.Possible distant elements:Vast green fieldsRice fieldsWheat fieldsRailway tracksSlow moving trainElectricity polesMain power transmission linesDirt roadsWater canalsRiver banksMango orchardsSmall villagesOpen countrysideHills in the distanceFlooded reflective fieldsTreesThe scene must feel huge and expansive.EXTREME DEPTH REQUIREMENTSThe camera must capture an exceptionally large portion of the scene.The visible area should extend far beyond the foreground.The scene should gradually blend into distant fields, roads, railway lines, canals, rivers, orchards, villages, trees, electricity poles and power lines.The image should reward zooming in with new details everywhere.The viewer should feel they are looking across dreams.Prioritize maximum visible distance while maintaining clarity and detail.HOMES DETAILSPre-1980 India/Pakistan only.Include:Flat rooftopsCharpaisClay housesCement housesLanternsKerosene lampsHand pumpsWater potsBicyclesBuffaloesCowsGoatsVillage treesNO MODERN ITEMS.NO MOBILE PHONES.NO MODERN BUILDINGS.NO LED LIGHTS.NO MODERN VEHICLES.HUMAN DISTRIBUTION (EXTREMELY IMPORTANT)The houses must feel fully populated and alive.Every visible rooftop must contain people.Avoid concentrating all people on a single rooftop.Human activity must be naturally distributed across the entire scene.Examples:Families sleeping on charpaisElderly people talking quietlyMothers sitting beside sleeping childrenChildren looking at the moonPeople drinking teaVillagers folding beddingSomeone reading under a lanternSomeone resting beside a water potSmall groups having quiet conversationsNo rooftop should appear abandoned.Even distant rooftops should contain tiny visible human silhouettes.The scene must feel rich with life.Target feeling:”The living houses under the night sky.”WINDStrong night breeze is mandatory.Visible effects:Tree leaves moving stronglyTree branches swayingCurtains flutteringBed sheets moving with airDupattas moving Hanging clothes swayingIMPORTANT:No dust.No sand.No reduced visibility.The air remains crystal clear.LIGHTINGCombination of:Cool blue moonlightWarm yellow lantern lightWarm house interior glowThe contrast between moonlight and lantern light must create a magical atmosphere.VISIBILITY REQUIREMENTSThe scene must remain bright enough to clearly see all rooftops, villagers, trees and distant landscapes despite being night.Avoid overly dark or low-visibility scenes.The environment should glow softly like a classic Studio Ghibli night illustration.COLOR PALETTEHighly vibrant and clearly visible colors.Rich blues.Deep turquoise.Emerald greens.Bright moonlit fields.Warm orange lanterns.Golden house lights.Luminous reflections.Highly saturated but natural colors.No dull colors.No muddy shadows.No excessively dark scenes.Every rooftop, tree, field and character should remain clearly visible.Dreamlike.Beautiful.Magical.ART STYLEStudio Ghibli inspired.Hand-painted animation look.Ultra detailed.Painterly textures.Soft cinematic shading.Storybook atmosphere.Rich environmental storytelling.FIRST TASKGenerate 10 unique video ideas.For each idea provide ONLY:TITLETWO-LINE STORY DESCRIPTIONDescription must be exactly two lines.No image prompts.No motion prompts.No explanations.IDEA RULESAll ideas must:Be night scenesBe peacefulBe relaxingHave strong windHave many villagersHave large environmental depthShow distant countrysideFeel nostalgicFeel aliveFeel beautifulWEATHER & BACKGROUND VARIATION (MANDATORY)Every idea must have a completely different night atmosphere and environmental background.Examples:Full moon over rice fieldsStar-filled sky with passing trainCloudy monsoon night beside a canalLight rain over village rooftopsnight in mango orchardsRiver reflecting moonlightFlooded fields after rainHeavy cloud night over distant villagesMoonlit countryside with power linesStarlit railway crossingNo two ideas should have the same weather and background combination.AFTER USER SELECTS AN IDEAGenerate:IMAGE PROMPTCreate an extremely detailed AI image generation prompt.Must include:Camera angleHuman countRooftop layouthouses layoutBackground depthDistant landscapeLightingCharacter activitiesColorsStudio Ghibli styleThe image prompt must create the FIRST FRAME of the video.MOTION PROMPT 01Create the first video clip.CAMERA RULE (EXTREMELY IMPORTANT)The camera is completely fixed and stationary.No camera movement of any kind.No zoom.No pan.No tilt.No glide.No push-in.No pull-out.No rotation.The entire video must feel like a locked-off tripod shot from a rooftop aerial viewpoint.Only environmental elements and characters move.Include:Tree movementLeaf movementCloth movementLantern flickeringCharacter micro-movementsEnvironmental movementNo dramatic actions.Everything remains peaceful.SOUND DESIGNOnly natural sounds:WindLeavesCricketsDistant dogsTrain soundsBirdsAnimal soundsLantern flameNO WIND EFFECTS.NO MUSIC.NO DIALOGUE.NO VOICEOVER.MOTION PROMPT 02Continue naturally from the last frame of Motion Prompt 01.CAMERA RULEExactly the same fixed and stationary camera.No camera movement of any kind.Add:New subtle movementsMore environmental lifeContinuity preservedSame peaceful atmosphereSOUND DESIGNNatural environmental sounds only.NO MUSIC.NO DIALOGUE.NO VOICEOVER.OUTPUT FORMAT10 IDEAS↓User Selects Idea↓IMAGE PROMPT↓MOTION PROMPT 01↓MOTION PROMPT 02↓Ready For Next IdeaOnce the AI returns your 10 ideas, reply with the number or title of the scene you want. It will then generate your image prompt and both motion prompts automatically. Take the image prompt straight into Google Flow (Step 3) and the motion prompts into the video generation step (Step 4).
Step 3: Create Your Image in Google Flow
Open Google Flow and go to the image generation section. Before you do anything else, set your aspect ratio to 9:16. This is non-negotiable. Vertical format is where these videos live YouTube Shorts, Reels, TikTok and generating in 16:9 and cropping later destroys composition and detail. Generate vertical from the start.
Paste your image prompt from Step 2 into the text field. Generate two or three variations and pick the strongest one. Don’t stop at the first result; AI image generation is probabilistic, and the second or third output often beats the first noticeably. Save your chosen image.
Step 4: Convert the Image to a Video Clip
Switch to Flow’s video generation mode. Attach your saved image and paste the motion prompt. Select your video model Flow currently offers options including OmniFlash and Veo 3.1, with Veo 3 pulling ahead on motion smoothness according to testing by creators in 2026. Generate your clip.
AI video generators typically produce 4 to 5 second clips. To build a 15 to 30 second Short, you need multiple clips chained together. Here’s the technique that makes those clips feel continuous rather than choppy: play your generated clip and pause on the very last frame. Screenshot that ending frame. Use it as your starting image for the next clip, paired with a new motion prompt that picks up where the last one left off.
This “chain linking” method creates the illusion of one continuous camera move. The viewer sees a scene evolving a train emerging from rain, moonlight shifting across rooftops rather than jarring hard cuts. It’s the single most important technique in this entire workflow for producing professional-looking results.
Editorial note: What stood out during research into this workflow was how much the chain linking step separates creators who get consistent results from those who don’t. Most beginner tutorials skip over it entirely, or mention it once without emphasizing it. But virtually every viral Ghibli-style channel does this you can often spot it in the subtle way the horizon line or rain angle stays perfectly consistent across a clip transition. It’s a small step that makes a disproportionately large difference.
Step 5: Repeat for Each Scene
Go back to your scene concepts from Step 2 and generate images and video clips for each one. If you’re building a 30-second Short, you’ll need roughly 6 to 8 clips. If you’re targeting a longer YouTube format, plan for 12 to 15. Keep your motion prompts varied: some clips can be pure environmental (wind through grass, rain on water), while others can include character movement (a figure walking, a child looking up at the sky).
Don’t aim for perfection on every clip. Generate, pick the best, move on. Over-polishing individual clips at this stage is where beginners lose hours. You’ll catch issues in the editing step.
Step 6: Assemble Your Clips in a Video Editor
Import everything into CapCut or InShot. Arrange your clips in the order that feels most narratively natural even if there’s no dialogue or story, there’s a visual flow to Ghibli-style content. The pacing usually moves from establishing wide shots toward intimate detail shots, then back out. Think of it like a slow exhale.
Add your audio at this stage. YouTube’s Audio Library has free, copyright-safe ambient and instrumental tracks that work well: rain ambiance, soft piano, acoustic guitar, or traditional Japanese instruments. Keep the audio quiet and atmospheric. These videos are not about the sound; the visuals carry the emotion.
Step 7: Apply Color Grading The Step That Separates Amateur from Viral
This is the step most beginners skip, and it’s the biggest mistake you can make. Raw AI video output tends to look flat, slightly washed out, and “computery.” Color grading is what makes it feel cinematic. In CapCut or InShot, open the manual color adjustment panel for each clip and apply these settings as a starting baseline:
Temperature: Push slightly warm (toward yellow/orange). This mimics the golden-hour and sunset lighting that Ghibli films use constantly.
Vibrance: Increase to around 60–65. Vibrance lifts muted colors without blowing out skin tones or skies.
Saturation: Moderate boost. You want greens to feel alive and skies to feel deep, not neon.
Contrast: Push to 50 or above. This adds depth and dimension the single most important color adjustment for this aesthetic.
Highlights: Gentle lift to brighten light sources like windows, the moon, or cloudy skies.
Sharpness: Slight increase for crisp detail on textures and edges.
Do a before/after comparison after grading your first clip. The difference is usually staggering. The ungraded version looks like AI output. The graded version looks like someone made a creative choice.
Is It Legal to Make Ghibli Style AI Videos?
This is a question worth addressing directly, because there’s real confusion online. Visual style is generally not protected by copyright the specific characters, scenes, and story elements from actual Ghibli films are protected, but the aesthetic (soft lighting, painterly backgrounds, a certain color palette) is not. As of 2026, U.S. and EU copyright frameworks treat generating “Ghibli-inspired” content as transformative work, not infringement, according to legal analysts covering the AI creative space.
READ MORE
The practical rule is simple: create original characters and scenes, not recreations of Totoro, Chihiro, or other specific Ghibli IP. Generate a girl watching fireflies, not a replica of a copyrighted scene. Also disclose that your content is AI-generated where platforms require it. YouTube, Instagram, and TikTok all have policies on this, and transparency protects your monetization eligibility.
It’s also worth noting that Studio Ghibli co-founder Hayao Miyazaki has been publicly skeptical of AI animation. That doesn’t make this workflow illegal, but it’s honest context to hold.
Common Mistakes to Avoid
After looking at how dozens of creators approach this workflow, the same errors come up repeatedly. The wrong aspect ratio is the most common: generating images in 16:9 and cropping to vertical is a composition killer. Always start in 9:16. The second is skipping the chain linking step in video generation, which produces jarring visual cuts that undercut the calm, flowing feel the aesthetic depends on. Third is neglecting color grading entirely, treating raw AI output as final. And fourth is overloading motion prompts asking for six simultaneous movements in one clip produces chaos. Pick two or three key motions and describe them precisely.
Read Also: How to Make an AI Underwater Documentary
One mistake that’s less obvious: expecting one perfect generation. AI tools are probabilistic. You’ll generate three images and pick the best. You’ll render two motion variations and choose the smoother one. Iteration isn’t failure; it’s the process.
Frequently Asked Questions
Can I make Ghibli style AI videos completely for free?
Yes. Google Flow handles both image and video generation at no cost on its free tier. ChatGPT’s free version handles prompt engineering. CapCut and InShot are free for mobile editing. The full workflow costs nothing to start, though free tiers have daily usage limits that slow high-volume production.
READ MORE
How long does it take to make one Ghibli style AI video?
A 15 to 30 second Short typically takes 2 to 4 hours the first time you do it, including prompt writing, generation, clip chaining, and color grading. With practice and a master prompt system in place, that drops to 45 to 90 minutes per video.
Do I need to credit Studio Ghibli when using this style?
No credit to Studio Ghibli is legally required when generating original scenes in a similar visual style. However, you should disclose that your content is AI-generated per platform policies, and you should never use actual copyrighted Ghibli characters or scenes as part of your output.
Which AI video tool produces the best Ghibli style results in 2026?
For motion fluidity, Veo 3 currently leads. For character consistency across multiple clips, Kling 3.0 is well-regarded. For a completely free starting point, Google Flow is the practical choice. The best results often come from generating images in one tool and converting to video in another, but this adds workflow complexity.
Why do my AI videos look flat and washed out compared to viral examples?
Almost always a color grading issue. Raw AI video output lacks the contrast and warmth that makes Ghibli-style content feel cinematic. Applying the manual adjustments described in Step 7 particularly contrast above 50 and a warm temperature shift resolves this in most cases.
Conclusion
Creating Ghibli style AI videos is genuinely accessible in 2026, and the workflow is more repeatable than most tutorials let on. The core steps reference screenshot, AI-generated prompts, image creation in Flow, chain-linked video clips, and manual color grading form a system you can run consistently rather than a one-off experiment. The creators getting millions of views aren’t doing anything technically beyond what’s described here. They’re just doing it consistently, with disciplined color grading and good prompt structure. Start with one video, learn the chain linking technique, and spend five minutes on color grading before you publish. The gap between “AI output” and “actual content people want to watch” is almost entirely in that last step.
