Wan

Modes: video · image · Models: 7

Vendor: Alibaba Model Studio (Wan) · Official API docs: Wan (Model Studio)

Wan 2.7 (by Alibaba) is a video model with text-to-video, image-to-video, reference-to-video, and video-edit workflows. It supports a start-frame keyframe, an optional driving-audio track, resolutions up to 1080p, and 5 / 10 / 15-second clips.

Models

id	Name	Input type
wan-2.7-t2v	Wan 2.7	t2v
wan-2.7-i2v	Wan 2.7 Image-to-Video	i2v
wan-2.7-r2v	Wan 2.7 Ref-to-Video	v2v
wan-2.7-video-edit	Wan 2.7 Video Edit	v2v
wan-2.6-t2v	Wan 2.6	t2v
wan-2.6-r2v	Wan 2.6 Ref-to-Video	v2v
wan-2.6-image	Wan 2.6 Image	t2i

CLI

# text-to-video
gen-ai generate -m wan-2.7-t2v \
  -p "a paper boat drifting down a rain-soaked street at night, neon reflections" \
  --ar 16:9 -r 1080P -d 10

# image-to-video from a start frame, with a driving audio track
gen-ai generate -m wan-2.7-i2v -p "the portrait slowly turns and smiles" \
  --start-frame ./portrait.jpg -a ./voice.mp3 -d 5

# edit an existing clip
gen-ai generate -m wan-2.7-video-edit -p "convert to hand-drawn anime style" --video ./clip.mp4

MCP

{ "name": "picsart_generate",
  "arguments": {
    "model": "wan-2.7-t2v",
    "prompt": "a paper boat drifting down a rain-soaked street at night, neon reflections",
    "aspectRatio": "16:9",
    "resolution": "1080P",
    "duration": 10
  } }

{ "name": "picsart_generate",
  "arguments": {
    "model": "wan-2.7-i2v",
    "prompt": "the portrait slowly turns and smiles",
    "startFrame": "./portrait.jpg",
    "duration": 5
  } }

Parameters

Full parameter surface for every model, sourced from gen-ai models info <id> --json. CLI flags show the primary short form; the canonical --kebab-case long form always works too.

wan-2.7-t2v — Wan 2.7

Try wan-2.7-t2v in Playground ↗

Input type: t2v

Param	CLI flag	Type	Values
prompt	-p	text	required
duration	-d	enum	5 · 10 · 15 (default 5)
resolution	-r	enum	720P · 1080P (default 720P)
aspectRatio	--ar	enum	16:9 · 9:16 · 1:1 · 4:3 · 3:4 (default 16:9)
negativePrompt	--neg	text	free text
enhancePrompt	--enhance-prompt	boolean	true · false (default true)
audioUrl	-a	file	audio
startFrame	--start-frame	file	image

wan-2.7-i2v — Wan 2.7 Image-to-Video

Try wan-2.7-i2v in Playground ↗

Input type: i2v

Param	CLI flag	Type	Values
prompt	-p	text	free text
duration	-d	enum	5 · 10 · 15 (default 5)
resolution	-r	enum	720P · 1080P (default 720P)
negativePrompt	--neg	text	free text
enhancePrompt	--enhance-prompt	boolean	true · false (default true)
startFrame	--start-frame	file	required image
endFrame	--end-frame	file	image
audioUrl	-a	file	audio

wan-2.7-r2v — Wan 2.7 Ref-to-Video

Try wan-2.7-r2v in Playground ↗

Input type: v2v

Param	CLI flag	Type	Values
prompt	-p	text	required
duration	-d	enum	5 · 10 (default 5)
resolution	-r	enum	720P · 1080P (default 720P)
aspectRatio	--ar	enum	16:9 · 9:16 · 1:1 · 4:3 · 3:4 (default 16:9)
negativePrompt	--neg	text	free text
imageUrls	-i	file	required image (up to 5)
videoUrl	--video	file	required video

wan-2.7-video-edit — Wan 2.7 Video Edit

Try wan-2.7-video-edit in Playground ↗

Input type: v2v

Param	CLI flag	Type	Values
prompt	-p	text	free text
resolution	-r	enum	720P · 1080P (default 720P)
aspectRatio	--ar	enum	16:9 · 9:16 · 1:1 · 4:3 · 3:4 (default 16:9)
negativePrompt	--neg	text	free text
videoUrl	--video	file	required video
imageUrls	-i	file	image (up to 3)

wan-2.6-t2v — Wan 2.6

Try wan-2.6-t2v in Playground ↗

Input type: t2v

Param	CLI flag	Type	Values
prompt	-p	text	required
duration	-d	enum	5 · 10 · 15 (default 5)
resolution	-r	enum	480p · 720p · 1080p (default 720p)
aspectRatio	--ar	enum	16:9 · 9:16 · 1:1 · 4:3 · 3:4 (default 16:9)
negativePrompt	--neg-prompt	text	free text
cfgScale	--cfg	range	1–10, step 0.5 (default 5)
startFrame	--start-frame	file	image

wan-2.6-r2v — Wan 2.6 Ref-to-Video

Try wan-2.6-r2v in Playground ↗

Input type: v2v

Param	CLI flag	Type	Values
prompt	-p	text	required
duration	-d	enum	5 · 10 (default 5)
resolution	-r	enum	720p · 1080p (default 720p)
videoUrl	--video	file	required video

wan-2.6-image — Wan 2.6 Image

Try wan-2.6-image in Playground ↗

Input type: t2i

Param	CLI flag	Type	Values
prompt	-p	text	required
count	-n	enum	1 · 2 · 4 · 6 · 8 · 10 (default 1)
negativePrompt	--neg-prompt	text	free text

Notes: wan-2.7-i2v keys off startFrame; the edit / reference variants take a videoUrl input.

Pricing

gen-ai pricing wan-2.7-t2v -d 10 -r 1080P

Cost scales with duration and resolution.