AIスタジオ 4 min read

AIホラー写真エディター

顔を失わずにセルフィーをホラーポートレートに変えます。

ローカルエディターアップロードなしプロンプトラボ

短い回答

短い回答

ローカルエディターと同一性ルールを備えた説得力のあるホラー肖像。

AI Horror Photo Editor (ja) editorial preview artwork
Girigo AI Studio AI Horror Photo Editor Local preview / copy-ready prompts / no upload

Local look lab

Preview the mood before you spend a credit

Upload a photo you own. The canvas, grade, grain, and caption stay in this browser.

No uploadNo accountPNG export

Choose a photo to start. Nothing leaves this browser.

Look

AIスタジオ

AIホラー写真エディター

多くのプロンプトは顔を忘れて失敗します。ここでは同一性、肌、ポーズを先に固定します。

ローカルエディターと同一性ルールを備えた説得力のあるホラー肖像。

AIスタジオ

進め方

  1. 編集する権利のある写真またはクリップを選びます。
  2. トリミング、色、構図をローカルエディターで確認します。
  3. プロンプトをコピーし、100%表示で顔を確認し、一度に一つだけ変更します。

Prompt lab

コピーできるプロンプト例

Gothic portrait

Midnight Braids

Best first test for a cheap identity-preserving edit.

Use the uploaded photo as the primary reference. Preserve the subject's recognizable facial identity, facial proportions, skin tone, approximate age, and natural anatomy. Transform the subject into an original gothic Halloween fashion portrait. Change the hairstyle into long violet braids with delicate silver hair accessories. Replace the clothing with an elegant black-and-purple velvet outfit with subtle lace and metallic details. Place the subject in a moonlit Victorian garden with soft fog, glowing lanterns, and dark autumn foliage. Use photorealistic editorial photography, realistic fabric texture, natural skin detail, and cinematic soft lighting. Keep the same number of people as the uploaded photo. Do not alter facial identity, add extra limbs, duplicate fingers, distort anatomy, or add text, logos, or watermarks.

Avoid: identity drift, plastic skin, extra fingers, extra limbs, text, logo, watermark

Open WaveSpeed
Quiet horror

Moonlit Widow

Use when you want atmosphere instead of costume spectacle.

Edit the uploaded portrait into a restrained Victorian ghost story still. Keep the subject's face, age, skin texture, body proportions, and pose recognizable. Add a sheer black veil, a high-collar charcoal dress, and a single silver brooch. Move the scene into an abandoned glass greenhouse at night, with wet stone, ferns, and a cold moonbeam cutting through the roof. Use soft volumetric light, realistic fabric, subtle film grain, and a desaturated blue-green palette. The mood should be melancholy, not gory. Keep one person only. Avoid a new face, changed ancestry, extra people, deformed hands, floating objects, captions, and watermarks.

Avoid: gore, new face, changed ethnicity, extra people, deformed hands, captions

Open fal
Analog horror

Static Saint

Good for testing how a model handles texture without losing identity.

Use the uploaded photo as the identity reference. Keep the subject recognizable while turning the portrait into an analog-horror broadcast still. Dress the subject in an ivory ceremonial coat with a cracked porcelain halo behind the head. Add restrained VHS scanlines, a faint timecode in the lower corner, and a cold fluorescent light from above. The scene is a small chapel with peeling paint and one red exit sign. Keep the face intact and the anatomy natural. Do not add a second face, a monster mouth, extra arms, readable brand text, or a watermark.

Avoid: monster face, second face, extra arms, brand text, watermark, heavy blur

Open Replicate

Model routing

どのモデルを使うか

Provider routing reference. Confirm current model names, availability, and pricing in each dashboard before a batch run.
ProviderBest forTypical routeDocs
WaveSpeedBudget-firstLow-cost image edits and image-to-video draftsSeedream edit, FLUX Kontext, Wan image-to-videoDocs
falSpeed-firstFast hosted queues and strong model coverageFLUX, Kling, and other hosted media modelsDocs
ReplicateExperiment-firstBroad catalogue for experiments and batch comparisonsCommunity and first-party image/video modelsDocs
SeeAPIRouting layerOne API surface for several model families and fallback routesImage and video inference endpointsDocs
KieCharacter-firstNano Banana image edits and Kling video routesNano Banana, Kling videoDocs

AIスタジオ

避けたいこと

English

英語の完全版には、失敗パターンの分析とモデル比較があります。

Read the English field note

Open the English full guide

FAQ

このページが素早く答えるべき質問

このページは写真をアップロードしますか?

いいえ。ローカルエディターはブラウザー内で動きます。プロバイダーにプロンプトを貼り付けた場合は、そのサービスの規約が適用されます。

商用利用できますか?

元写真の権利があり、モデルのライセンスが商用利用を許し、必要な場所でAI利用を明示する場合に限ります。他人の顔は同意なく使わないでください。

確認済みソース

このページの根拠となる公式情報