Skip to main content
POST
cURL
muse-video-i2v animates a still image. Pass one image and it becomes the first frame; pass a second image and Muse Video interpolates a shot from the first frame to that last frame, giving you exact control over where the motion begins and ends.

Model at a glance

Parameters

Reference images supplied through reference_images are rejected here — that alias belongs to muse-video-r2v, which treats every image as a subject reference rather than a frame.

First frame only vs. first and last

First-to-last is the reliable way to land on an exact composition. Describe the journey in the prompt rather than both endpoints — the model already knows the endpoints.

Workflow

  1. POST /v1/videos/generations with first_frame_image (and optionally a second entry in image_urls).
  2. Read data.task_id.
  3. Poll GET /v1/videos/status/{task_id} until completed.
  4. Download from data.result.

Image requirements

  • Use plain, public HTTPS URLs. Signed links that expire, hotlink-protected hosts, and URLs that require a login all fail at submission.
  • Portrait, product and landscape photos all work; keep the subject unobstructed in the first frame.
  • For real-person portraits, prefer wide, well-lit shots — extreme close-ups give the model very little to animate.

Cost

Billing is per output second and, unlike reference-to-video, the frames you supply do not add to the bill: duration × per-second price.

Authorizations

Authorization
string
header
required

Use your APIPod API key as a Bearer token in the Authorization header.

Body

application/json
model
string
required

Public APIPod model ID.

Allowed value: "muse-video-i2v"
prompt
string
required

Describe the video to generate. Required, up to 4000 characters.

Required string length: 1 - 4000
first_frame_image
string
required

Required. Public URL of the first frame image. Pass a second URL in image_urls (up to two) to use it as the last frame.

duration
enum<integer>
default:5

Video duration in seconds.

Available options:
5,
10,
20,
30,
40,
50,
60
aspect_ratio
enum<string>
default:16:9

Aspect ratio of the generated video.

Available options:
16:9,
9:16,
1:1,
4:3,
3:4
resolution
enum<string>
default:720p

Output resolution. MU Video currently supports 720p only.

Available options:
720p
style_preset
enum<string>
default:none

Visual style preset applied to the video.

Available options:
none,
cinematic_realistic,
2d_anime,
3d_xianxia,
american_comic,
cyberpunk_neon,
ecommerce_ugc
generate_audio
boolean
default:true

Whether to generate voice and sound. Set to false to mute voices.

Response

200 - application/json

Task accepted

Standard response for task submission

code
integer
required
data
Task Submit Data · object
required

Data content for task submission

message
string
required