作例リール
実際の生成結果です。各クリップの下にあるのが、それを生み出したプロンプトです。
A lone rider crosses a mirror-flat salt lake at dawn, perfect reflection below, wide anamorphic framing, hooves echoing in the silence
Veo 3.1A row of tiny robots on an assembly line, each painting one frame of a film strip; the strip runs off the line and comes alive as a moving picture
Seedance 2.0A sports drink bottle bursts through a wall of ice, shards flying in slow motion, electric blue splash forming a wave behind it, punchy commercial energy
Kling 2.6A corgi wearing tiny sunglasses rides a skateboard down a beach boardwalk, tongue out, upbeat summer energy
Veo 3.1 FastCold brew coffee pours over clear ice in extreme slow motion, amber cascades and caramel swirls, deep rich sound design, premium commercial
Veo 3.1The wind-up fox toy comes to life: gears whir, it blinks, stretches, and trots forward across the workbench, warm lamp light
Seedance 2.0POV walking through a neon night market, quick pans across sizzling food stalls, steam and lantern light, energetic vlog style
Kling 2.6A bakery's morning in time-lapse: dough rising, croissants blooming golden in the oven, steam curling, the open sign flipping
Veo 3.1 FastAn old lighthouse keeper lights the great lamp as a storm arrives, rain streaking the glass, a small boat's light making it safely home below
Veo 3.1The same emerald wind-up fox toy walks carefully along a moonlit bookshelf, past leaning books, tail swaying with each mechanical step
Seedance 2.0Origami birds fold themselves out of sheet music and fly through a dark concert hall, camera chasing them through a spotlight
Kling 2.6A pencil storyboard sketch of a rocket launch comes alive, the drawing transitioning to a full cinematic render mid-liftoff
Veo 3.1 FastA ballerina made of golden autumn leaves forms mid-spin and dances, leaves scattering with each turn on a black stage, wind audible
Veo 3.1Glowing data streams flow between dark server racks and weave themselves into a strip of moving film frames, abstract and precise
Seedance 2.0A smartwatch rotating in studio light, screen cycling through watch faces, macro detail of the crown clicking, minimal white set
Kling 2.6A ceramic mug on a colored backdrop spins and snaps through four colorways in quick cuts, product content at scale
Veo 3.1 Fast
各モデルが入力できるもの
- Seedance 2.5Text, images, a reference video, a reference audio clip, first and last frameNative audio4–15 s
- Seedance 2.0Text, images, a reference video, a reference audio clipNative audio4–12 s
Gemini Omni FlashText, an imageNative audio3–10 s- Veo 3.1Text, reference images, first and last frameNative audio4, 6, 8 s
- Veo 3.1 FastText, an imageNative audio4, 6, 8 s
- Kling 2.6Text, an imageNative audio5–10 s
- Veo 2Text, an imageSilent5, 6, 8 s
よくある質問
- Do I need a plan to generate video?
- Yes. Clips are generated in Studio with video credits from your plan, and each clip's cost depends on the model. Starter includes 10 video credits a month; Expert includes 35.
- Which AI video generator is the best right now?
- It depends on the shot. Veo 3.1 gives the most film-like single shot with speech; Seedance 2.0 is the one to use when a character or product must look the same across shots; Kling 2.6 wins on motion. The table above lists what each takes in.
- Can I turn a photo into a video?
- Every model on this page accepts an image. Upload the still, describe the motion in one sentence, and pick a model; Seedance 2.0 also takes a reference video and audio clip.
- Do the videos have sound?
- Seedance 2.0, Veo 3.1, Veo 3.1 Fast, Veo 3 Fast and Kling 2.6 render native audio with the footage. Veo 2 and the original Seedance produce silent clips.
- How long can a clip be?
- It varies by model: Seedance 2.0 takes 4 to 12 seconds, Kling 2.6 five to ten, and the Veo family 4, 6 or 8 seconds. Longer pieces are made by generating shots and cutting them together.
- What resolution do I get?
- Seedance 2.0 and the Veo 3.1 family output 1080p in Studio. Download the file from your library.
- Can I use the clips commercially?
- Clips you generate are yours to use under the terms of your plan and the model provider's policy. Check the provider's page for any restriction on faces or brands you did not create.
- Is there an API for video?
- Yes. The same model ids are served by the NinjaChat API as async jobs with polling or webhooks. Each developer page shows the request shape and the live per-generation price.