HappyHorse 1.1 Reference-to-Video

Alibaba HappyHorse 1.1 reference-to-video. Generates a 1080p clip anchored to up to 9 reference images for character/scene consistency, with synchronized native audio and multilingual lip-sync. Use @character1, @character2 tags in the prompt to bind subjects.

PRICINGstarting $0.14/request

LATENCY~220 seconds average

RESOLUTION720P/1080P

BEST FORvideo, reference-to-video, alibaba

Parameters

PROMPT*

REFERENCE URLS*

0/9 items

ASPECT RATIO

RESOLUTION

DURATION

SEED

Estimated Cost

Base cost per second$0.14

Total$0.14

Pay only for what you use. No minimums.

Output Preview

Ready

No output yet

Run the model to see generated results.

API Example— Current Parameters

generate.py

import requests

result = requests.post(
    'https://api.muvi.video/v1/jobs/submit',
    headers={
        'Authorization': f'Bearer {API_KEY}',
        'Content-Type': 'application/json'
    },
    json={
  'model': 'alibaba/happyhorse-1-1/reference-to-video',
  'input': {}
}
)

Get Job— Poll for result

get_job.py

import time

job_id = result.json()['jobId']

while True:
    response = requests.get(
        f'https://api.muvi.video/v1/jobs/{'{job_id}'}',
        headers={'Authorization': f'Bearer {'{API_KEY}'}'}
    )
    data = response.json()['data']

    if data['status'] == 'completed':
        print('Done!', data['outputs'][0]['url'])
        break
    elif data['status'] == 'failed':
        raise Exception(f"Job failed: {'{'}data['error']['message']{'}'}")

    time.sleep(2)

HappyHorse 1.1 Reference-to-Video - AI Model | Muvi