AI生成画像とは何ですか?
AI 生成画像は、テキスト、画像、または混合入力から AI モデルを使用して作成または変換されたビジュアルです。
実際に生成されたAI動画や画像の作例をまとめて確認できます。ワークフローごとの出力傾向を見比べながら、近い方向性を再現するための参考にできます。
このページでは、実際のAIクリエイションを通して、各ツールやモデルでどんな表現が作れるのかを確認できます。
Cinematic live-action shot of a man sitting at a white kitchen island in a modern home, eating spaghetti. Natural hand movement, realistic chewing and swallowing, medium close- up, subtle push-in camera movement, daylight from a side window, grounded household details, 24fps realism.
First-person GoPro view jogging through misty redwood forest at dawn, as playful Bigfoot runs beside the camera waving; hyper-real foliage, soft god-rays, warm morning birdsong, light footstep foley; 8-second continuous take, 4-K, 30 fps.
There is no sound in the video. All characters are humanized chess pieces doing human activities. The animation should be in the same pop art graffiti style as the images. Begin by fading into the rap concert. focus on the rapper on stage. then pan the camera around to see the crowd, then change camera angle to view from the back of the stadium to the stage and the rapper rapping.
2x2 collage. Keep subject and outfit the same. Slightly modify pose and facial expression. Frame 1: shy. Frame 2: seductive lip bite. Frame 3: confident. Frame 4: longing face with tongue out. no text { "subject": { "description": "20yo cute Korean idol, bare shoulders, posing in marble bathroom.", "mirror_rules": "null", "age": "20", "expression": { "eyes": { "look": "direct", "energy": "innocent", "direction": "at lens" }, "mouth": { "position": "slightly parted", "energy": "soft" }, "overall": "cute and alluring" }, "face": { "preserve_original": false, "makeup": "Korean celebrity styling, dewy, fair, youthful" }, "hair": { "color": "dark brown", "style": "messy updo", "effect": "stray strands" }, "body": { "frame": "slim", "waist": "hidden", "chest": "covered by arms", "legs": "hidden", "skin": { "visible_areas": "face, shoulders, chest, arms", "tone": "pale white", "texture": "feels smooth, soft like porcelain", "lighting_effect": "subtle highlights" } }, "pose": { "position": "standing", "base": "lower edge", "overall": "arms crossed over chest" }, "clothing": { "top": { "type": "bare shoulders", "color": "null", "details": "null", "effect": "null" }, "bottom": { "type": "hidden", "color": "null", "details": "null" } } }, "accessories": { "headwear": "null", "jewelry": "diamond studs, delicate pendant, layered diamond ruby bracelet", "device": "null", "prop": "null" }, "photography": { "camera_style": "smartphone selfie aesthetic", "angle": "slightly high", "shot_type": "bust shot", "aspect_ratio": "3:4", "texture": "soft digital noise, slight blur", "lighting": "off-center overhead, soft bounce", "depth_of_field": "shallow" }, "background": { "setting": "bathroom", "wall_color": "grey white marble", "elements": [ "bathtub edge", "wall sign" ], "atmosphere": "private", "lighting": "ambient" }, "the_vibe": { "energy": "relaxed", "mood": "intimate", "aesthetic": "clean", "authenticity": "candid", "intimacy": "high", "story": "late night unwinding", "caption_energy": "night routines" }, "constraints": { "must_keep": [ "jewelry", "marble wall" ], "avoid": [ "visible clothing", "heavy makeup" ] }, "negative_prompt": [ "old", "ugly", "heavy shadows", "clothes", "deformed", "CGI" ] }
An extreme metal band performs live, and the demons dance wildly
Make the people in the 3x3 grid of images move one after another
Comedic selfie skit ends on a hard freeze-frame as the protagonist raises a middle finger to the off-screen “haters”; record-scratch stops the music, bold caption POPS in sync; seamless 8 s loop, 4-K, 30 fps.
Epic aerial chase: emerald dragon twists between floating basalt spires above storm-lit ocean, breathing swirling blue fire; camera dives alongside, slow-motion wing beats send shock-waves through clouds; thunderous orchestral score, 8 s, 4-K
{ "SceneType": "Hyper-realistic, surreal composite commercial photograph", "Perspective": "Set on a wooden tabletop", "MainElement": { "Object": "A large, oversized smartphone", "Placement": "Dominates the frame", "Effect": "Creates a strong dimensional-break illusion" }, "InsideTheScreen": { "Subject": { "Reference": "Man from image_0.png", "Appearance": "Appears much larger than normal", "Environment": "Close-up winter environment", "Pose": "Smiling happily, holding a clear glass with a visible Pepsi logo on it" } }, "OutsideTheScreen": { "InteractingHand": "A real human hand", "Action": "Pours dark Pepsi cola from a cold, condensation-covered bottle directly toward the phone screen", "SurrealEffect": "The liquid visually breaks through the display and flows into the glass the man is holding inside the screen" }, "SupportingObjects": [ { "Object": "Closed book", "Placement": "On the wooden table beside the phone" }, { "Object": "Black pen", "Placement": "On the wooden table beside the phone" } ], "StyleAndAtmosphere": { "Lighting": "Warm natural lighting, soft cinematic shadows", "Details": [ "Ultra-sharp liquid splash details", "Hyper-realistic skin texture", "Realistic reflections on the glass and the phone screen" ], "Aesthetic": [ "Surreal cinematic advertising look", "Professional commercial photography quality", "8K realism" ] } }
スタイル、被写体、変換タイプごとに整理された作例を見ながら、入力やモデル選択が結果にどう影響するかを比較できます。
シンプルなプロンプトや画像入力から、フォトリアルな写真、スタイライズされたアート、商品ビジュアル、ポートレート、変換系の表現まで幅広く生成できます。ここに並ぶ事例はすべて実際の出力なので、ワークフローの実力を具体的に把握できます。
スタイル、被写体、変換タイプごとに整理された作例を見ながら、入力やモデル選択が結果にどう影響するかを比較できます。
気になる表現を見つけたら、対応するワークフローへそのまま移動して再現できます。近いプロンプトや参照画像、モデル選択を使えば、似た方向性の結果を素早く試せます。
よくある質問
AI 生成画像は、テキスト、画像、または混合入力から AI モデルを使用して作成または変換されたビジュアルです。
はい。 このページの多くの例は、リンクされたツールや同様の設定を使用して再作成できます。
リアルな写真、様式化されたアート、変換、前後の例などを探索できます。
いいえ、ほとんどのワークフローは初心者にとって十分シンプルになるように設計されています。
次のステップ
一致するワークフローとモデル ページを検討し、同じ視覚的な方向性を念頭に置いて、インスピレーションから生成に移行します。