FlyAIgh
首页/模型/WAN 2.7 Video
by Alibaba

WAN 2.7 Video AI 视频生成

Alibaba's WAN 2.7 video model — multi-shot narrative generation with full mode coverage (t2v / i2v / r2v / edit). Up to 1080p, 2–15s clips.

WAN 2.7 Video45 cr/5s

FlyAIgh 是一个独立平台,为 WAN 2.7 Video 及其他 AI 模型提供统一调用入口。我们与 Alibaba 不存在任何关联、背书或赞助关系。

WAN 2.7 Video 各版本与参数

参数WAN 2.7 Video
时长2–15s
画面比例16:9 · 9:16 · 1:1 · 4:3 · 3:4
分辨率1080p
原生音频
图生视频
参考图生视频
积分/秒9
5 秒片段消耗45 cr/5s

什么是 WAN 2.7 Video?

WAN 2.7 Video is Alibaba Cloud's flagship text-to-video model and the most versatile single model on FlyAIgh in terms of mode coverage. It supports all four input modes — text-to-video, image-to-video (first or last frame), reference-to-video (up to 4 reference images) and video editing (modify an existing clip with a text instruction) — under one model id, with the platform routing automatically based on what you provide.

The headline capability is multi-shot narrative. Where most video models produce a single continuous shot, WAN 2.7 can compose multiple shots within a single generation — transitioning between camera angles, locations or story beats. This makes it particularly strong for ad concepts, short-form storytelling and scene-driven content where a single static framing isn't enough.

Clips run from 2 to 15 seconds in 1-second increments across five aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4). Output resolution is 720p or 1080p — pick 1080p for hero shots and final deliverables. Pricing is 8 credits per second — 40 credits for a standard 5-second clip, 80 for 10 seconds.

WAN 2.7 Video sits in the same family as WAN 2.7 Pro (the image-side flagship). Both share Alibaba's underlying generation stack but specialize for their respective modalities. The video model adds a few unique parameters: `prompt_extend` (the model rewrites your prompt for better narrative coherence) and `audio_url` (drive the video to a supplied audio track for music-synced output, on the text-to-video path).

WAN 2.7 Video is particularly strong at multi-shot composition, narrative pacing and Chinese-cultural visual aesthetics. It is weaker than VEO 3.1 Quality for ultra-fine textural detail and weaker than Sora 2 Pro for photorealistic faces. Pick WAN 2.7 Video when you need narrative breadth or full mode coverage in a single generation; pick the specialists when you need maximum fidelity in a single shot.

On FlyAIgh, WAN 2.7 Video runs through Alibaba Cloud Bailian with no separate Alibaba account, API key or Mainland China registration required.

WAN 2.7 Video vs Kling V3 vs HappyHorse 1.0

能力WAN 2.7 VideoKling V3HappyHorse 1.0
原生音频
图生视频
参考图生视频
首末帧

WAN 2.7 Video 常见问题

What is WAN 2.7 Video?+
WAN 2.7 Video is Alibaba Cloud Bailian's flagship video generation model. It's the most versatile single model on FlyAIgh for mode coverage — supports text-to-video, image-to-video, reference-to-video and video editing in one model id, with multi-shot narrative composition as the headline feature.
How much does WAN 2.7 Video cost on FlyAIgh?+
WAN 2.7 Video is 8 credits per second — 40 credits for a 5-second clip, 80 credits for 10 seconds. Same per-second rate as Kling V3 and Vidu Q3 Pro.
What is multi-shot narrative?+
WAN 2.7 can compose multiple shots within a single generation — transitioning between camera angles, locations or story beats — where most other video models produce a single continuous shot. This is particularly useful for ad concepts and short-form storytelling that needs more than one framing.
What resolutions and durations does WAN 2.7 Video support?+
WAN 2.7 outputs at 720p or 1080p across five aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4) with durations from 2 to 15 seconds in 1-second increments. The wider duration range and 1080p option set it apart from VEO 3.1 (fixed 8s) and most other models on FlyAIgh.
Is WAN 2.7 Video related to WAN 2.7 Pro?+
Yes — WAN 2.7 Video and WAN 2.7 Pro are sibling models in Alibaba's WAN 2.7 family. WAN 2.7 Pro is the image-side flagship (4K text-to-image with Thinking Mode); WAN 2.7 Video is the video-side flagship. Both share the underlying generation stack and aesthetic, specialized for their respective modalities.
还有未覆盖的问题? 联系客服 →

在 FlyAIgh 上使用 WAN 2.7 Video

一个积分钱包,所有顶尖 AI 模型。免费注册,无需信用卡。

开始生成