Pay only when you press Make · Stripe or PayPal · Failed video = credits back
By uploading a photo you confirm you're 18+ and own it. · Rules
▶ Real result14 s · 16:9
Genjutsu AI
1 photo → 14 s 16:9 HD video, ready in about 3–6 min
30 credits
The short answer
Effect Farm makes Genjutsu AI from 1 photo. Genjutsu AI puts you in the viral crowd scene: you stand still in a white shirt while rows of identical figures bow around you. You get 14 s at 720p, 16:9, MP4. Template soundtrack included. One video costs $9.99. Watch the samples and see your price before you pay. No sign-up needed: pay as a guest with card or PayPal. Usually ready in 3–6 minutes. A failed video returns its credits.
The example shows the full scene from one photo. You stand in the middle of rows of identical uniformed figures in a stone square. For nine seconds nobody moves. Then the whole crowd bows at once and you stay upright. Play it with sound to hear the hit on the bow.
Watch the samples, see your price, then pay. No subscription. Pricing from $1.42 per video. A failed or blocked video returns its credits automatically. Unused credits can be refunded within 14 days.
Who runs Effect Farm
Run by Effect Farm Team, an independent online service.
Payments by Stripe and PayPal. You can pay as a guest.
Every paid video: 1 remake within 7 days, any reason. Refund policy
Updated
More about Genjutsu
What is the Genjutsu trend?
AI versions of a crowd scene spread on TikTok in September 2026. A man in a white shirt stands still among rows of identical uniformed people. Then the whole crowd bows deeply at once and he stays upright. A clip posted on 17 September 2026 had 781,600 views by 12 October.
The clip works because of contrast. One still figure stands among rows of identical people who suddenly bow together. Our version rebuilds that scene around your photo: the rows, the white shirt, the pause and the bow.
Watch the examples to check the movement and framing, then add the photos requested by the tool: 1 photo You upload still photos. You do not need to record a performance or write a prompt. The sample photo inset shows the input used for that example, so you can compare the starting image with the finished video.
2. Check your selection and make it
Check the selected effect and the price before paying. Guest checkout is available. Sign-in is optional. This effect uses 30 credits per video.
3. Open your link and download
Generation usually takes 3–6 minutes. You can stay on the page or return through the private link in your email. Watch the complete result and check the face, hands and movement before sharing. A failed generation returns its credits. Not happy with a delivered result? It includes 1 remake within 7 days, or write to support@effectfarm.com.
Use one clear photo of yourself with your face easy to see. A photo from the waist up or a full-body photo both work. You do not need a white shirt: the AI dresses you in one. Use a photo with only you in it. Uploading and checkout work without signing in.
Use JPG, PNG, WebP, HEIC, up to 30 MB. Photos are resized before upload. You can swap the picture before you start. You must be 18 or older. Confirm photo rights and adult age. We delete uploads after 7 days and never use them to train models.
You get
MP4, 720p, 16:9, 14 s. Template soundtrack included.
You get a 14-second horizontal (16:9) video of the Genjutsu crowd scene. You stand still in a white shirt among rows of identical figures in sand-coloured uniforms. At about nine seconds the crowd bows deeply together, and you are the only one left standing. The video has our own dramatic soundtrack. The figures are generated and fictional. Open your private result link to watch and download it.
This effect uses 30 credits for one video. Pack pricing starts from $1.42 per video. Pack credits never expire. Compare packs before buying more credits.
Photo tips for a better Genjutsu video
A good input makes the subject easier to read. Open the photo at its original size and check the important details before uploading. A bright, sharp photo is more useful than a heavily edited one. Match the crop to the effect: a face-focused scene and a routine with full-body steps need different framing.
Use even light and keep your face clear.
Match the framing of the dance. Show your feet for full-body moves.
Choose a simple background with only you in the photo.
Use a sharp original photo without heavy filters.
If a face is partly hidden, choose another photo instead of expecting the model to fill it in accurately. For a pet slot, make sure there is only one animal and its face is visible. Avoid collages: each slot is for its own still image. The samples provide a useful check on how much of the subject to include.
Check the finished result
Movement can change fine details in the face, hands, clothing or background. Play the whole result before downloading or sharing it. Not happy with a delivered video? It includes 1 remake within 7 days, with the same photo or a new one, or write to support@effectfarm.com. A generation that fails returns its credits automatically.
Genjutsu AI video ideas
Use these as starting points for your photo and for the finished clip. They are ways to use the effect, not extra controls. The template sets the movement.
The still one in the crowd
The trend is a contrast: rows of identical people, one person who does not fit, and a sudden bow. Your video keeps that structure. Use a photo with a calm, confident face, because you look at the camera while everyone else bows.
How this differs from Higgsfield Genjutsu
Higgsfield Genjutsu is a feature inside Higgsfield that rebuilds part of an existing video. Its workflow starts with a source video of 4–30 seconds and accepts up to 30 optional reference images. Effect Farm is not affiliated with Higgsfield.
One still photo, no video upload
Here, you supply one still photo of yourself. You never upload a video. The scene, the crowd, the white shirt and the bow are fixed, so there is nothing to trim or arrange.
Want to dance instead?
If you want a solo dance from your photo, use Make Picture Dance. It has four preset routines: Hip-hop, K-pop, Step & clap and Arms up.
You can watch the samples and see your price before you pay, with no sign-up. Each video you make is paid: one video costs $9.99. Pack credits never expire.
Do I need to sign up?
No. Pay as a guest with card or PayPal. Upload without an account and see the price before generation. Your payment provider collects your email. Sign-in is optional. Keep your private order and wallet links to return to your videos and credits.
What if my video fails? Can I get a refund?
A failed video returns its credits to your wallet automatically. This is a credit return, not an automatic cash refund. For unused-credit refunds, email support@effectfarm.com from your checkout email within 14 days of purchase.
Do I need an account?
Upload your photo, see the price and pay as a guest. No sign-up needed. Pack credits never expire and are tied to your private wallet link (or your account if you sign in). Guest purchases open through their private links.
What happens in the Genjutsu AI video?
You stand still in a white shirt in the middle of rows of identical uniformed figures. Nobody moves for about nine seconds. Then the whole crowd bows deeply at once, and you stay upright and look at the camera.
How long is a Genjutsu AI video?
14 seconds, horizontal (16:9), with our own soundtrack.
What kind of photo works best?
One clear photo of yourself in good light, facing the camera, with only you in it. You do not need to wear a white shirt; the AI adds it.
Are the people in the crowd real?
No. The crowd is generated. The figures are fictional adults in uniform, not real people. Only you come from a photo.
How is the video made?
The AI makes a new video of you from your photo. The camera, the pause and the bow follow the trend clip. It does not edit the original clip.
What if the result looks wrong?
A failed video gives its credit back automatically. If you are not happy with a delivered video, it includes 1 remake within 7 days of delivery, with the same photo or a new one.
More effects
Compare the sample, photo requirements and output before choosing another effect.