4つのモデルで、最大2K・4〜30秒の動画を作れます。そのうち4つは音声付きです。料金は1本25クレジット($0.25)から。サブスクなしで、クレジットに有効期限はありません。 Artificial Analysis Text to Video Arena(2026年10月30日確認)では、Wan 3.0が1位(Elo 1,175)、MiniMax H3が3位、Seedance 2.5が4位です。
動画の料金は720p・音声付き(MiniMax H3: 768p)で、各モデルのいちばん安いバージョンの価格です。解像度を上げると料金も上がります。 Arena順位は、ブラインドの1対1比較の投票による順位とEloスコアです(AA-Video-T2V v2.1, with audio、2026年10月30日確認)。出典: Artificial Analysis Text to Video Arena
“Slow dolly-in on a matte terracotta vase standing on a travertine plinth in a sunlit plaster room. Late-afternoon light slides slowly across the wall as dust drifts through the beam. Calm, premium product film, shallow depth of field. Audio: soft room tone and a faint breeze.”
“A strawberry sandwich cookie slowly completes one full rotation while the camera moves smoothly around it and keeps the product centered.”
ヒント:被写体は画面の中央に。字幕やアプリのボタンに隠れにくくなります。
360クレジット
MiniMax H3
カメラに向かってセリフを言う
レンズを見て、書いたとおりのセリフを言い、周りには厨房の音。高速版のH3 Max Turboで生成しました。
テキストから動画
768p
5秒
音声付き
16:9
“A chef in her forties looks straight into the camera in a bright open kitchen and says, "The secret is you never crowd the pan." She smiles and turns back to the stove. Clean daylight from a window, white tile, flour on her apron, a strand of hair loose. Medium close-up, 50mm, slight handheld. Sound: her voice close and clear, a gentle sizzle, a knife on a board off screen.”
ヒント:セリフは引用符で囲み、誰がどう言うかを書きます(「looks straight into the camera … and says, “…”」)。
“One unbroken handheld take, direct-flash 35mm snapshot look, fine grain, white tile and bright window light. 0-5s: from behind, a young pastry cook in a white jacket and a long black apron pipes the last pale-yellow buttercream flowers onto a small white cake on a scrubbed wooden bench in a narrow tiled pantry, steel bowls stacked on the shelves. 5-10s: she puts the piping bag down, turns the cake once on its plate to check it from every side, and straightens one flower with a fingertip. 10-15s: she lifts the cake on its plate, turns and carries it through the cramped kitchen past hanging ladles and a steaming stockpot, the camera following at shoulder height. 15-20s: she shoulders open a back door into a quiet, sunlit courtyard and walks across the flagstones. 20-25s: she sets the cake down on a weathered table laid for one and steps back to look at it. 25-30s: she wipes her hands on the apron and walks back inside, leaving the cake alone on the table in the sun. Sound: piping bag squeeze, a pot lid rattling, the door creak, birds outside.”
“Real-time speed, 35mm documentary look, flat daylight from tall paper windows in an old wooden dojo with straw-yellow tatami. Two women judoka, one in a white judogi and one in a deep blue judogi, grip each other's lapels and sleeves. The woman in white steps in, turns her hip under her opponent and throws her in a clean hip throw; the woman in blue rotates over her hip through the air, lands flat on her back with a loud slap of the mat and immediately rolls up to her knees. Camera: low handheld, side-on, following the throw. Sound: bare feet on tatami, a sharp exhale, the slap of the landing.”
ヒント:誰が何をするかを服装で区別します(「the woman in white … the woman in blue」)。モデルが2人を取り違えにくくなります。
“The theme is "mechanical flower bloom". The image needs to highlight the generative video model's strengths in lighting and shadow, art detail, realism, camera movement, and cinematic character feel. The overall style is a high-end tech-brand ad with strong visual impact. The video uses a single-take macro push-in shot, starting from a metal flower bud in darkness, gradually entering the precision mechanical structure inside the petals, and finally climaxing on the image of the mechanical flower fully blooming with light radiating outward. It requires realistic physical lighting, refined metal and glass materials, delicate mechanical motion, stable composition, cinematic color grading, no subtitles.”
“Medium close-up, 50mm. A night-market noodle vendor in a white apron stands behind a steaming wok under bare hanging bulbs. He looks into the camera and says: "Thirty years. Same pan, same recipe." Then he tosses the noodles high and flames jump up from the wok. Camera: gentle living handheld that barely breathes, slight push-in. Audio: his voice, sizzling oil, crowd chatter, distant scooters. Photorealistic, warm practical light.”
ヒント:セリフを先に、大きな動きをそのあとに置きます(「first …, only then …」)。
“Stop-motion claymation: a round green frog in a knitted scarf carries a stack of pancakes across a tiny kitchen set, sets it down, then looks at camera and blinks. Visible fingerprints in the clay, felt walls, warm practical lamps, stepped 12fps motion. Sound: playful pizzicato strings, soft clay squeaks, a plate clink.”
Wan 3.0(アリババの動画生成AI)は、プロンプトにいちばん忠実です。Kreaのテストでは「the most literal, dramatic version of a prompt」を出し、Fuserが同じ入力で行った5回の比較では、Seedance 2.5が見落とした細部を拾いました。1回で最長30秒まで作れ、1秒あたりの料金はSeedance 2.5より安くなります。弱点は、途中で別アングルに切り替わることがある点と、音が大きめな点です(Krea、Fuser)。
MiniMax H3(Hailuo AI(ハイルオAI)の動画モデル)は、ここでは2つの版があります。H3 Max Turboはこのページでいちばん速く安い動画で、本番にクレジットを使う前に言い回しを試すのに向いています。通常版のH3は2Kにも対応します。768pは720pではなく1344×768で出力され、短いプロンプトは書き換えられるため、カットが入ることがあります(Krea)。
プロンプトは、zzo.aiでは翻訳せず入力したまま各モデルに送ります。日本語のプロンプトでどこまで意図どおりになるかは、まだ十分に確かめていません。このページのサンプルはすべて英語のプロンプトで作っており、思ったとおりにならないときは英語で書くと安定する場合があります。セリフの言語は「says in English」のようにプロンプトで指定します。