在WAN 2.2上的Stable Video Infinity 2.0 Pro:将静态图像扩展为理论上无限长的视频,同时保持一致的角色ID.
VITA-Group / EPFL视频生成发布日期 Dec 26, 2025原生推理在WAN 2.2上的Stable Video Infinity 2.0 Pro:将静态图像扩展为理论上无限长的视频,同时保持一致的角色ID.
也称为 VITA-Group / EPFL SVI 2.0 Pro, SVI-2.0-Pro, svi-2-0-pro
svi-2-0-proPOST /v1/videos/generationsLive pay-as-you-go rates from the EmpirioLabs catalog. You are billed only for what you use, with no monthly minimum.
SVI 2.0 Pro runs through POST /v1/videos/generations. The request returns a job_id right away; poll GET /v1/jobs/{job_id} until the job completes and read the output URLs from the result. Get an API key from the EmpirioLabs dashboard.
curl https://api.empiriolabs.ai/v1/videos/generations \
-H "Authorization: Bearer $EMPIRIOLABS_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "svi-2-0-pro",
"prompt": "Describe what you want SVI 2.0 Pro to generate."
}'curl https://api.empiriolabs.ai/v1/jobs/JOB_ID \
-H "Authorization: Bearer $EMPIRIOLABS_API_KEY"import requests
response = requests.post(
"https://api.empiriolabs.ai/v1/videos/generations",
headers={"Authorization": "Bearer YOUR_EMPIRIOLABS_API_KEY"},
json={
"model": "svi-2-0-pro",
"prompt": "Describe what you want SVI 2.0 Pro to generate.",
},
)
job = response.json()
# Generation runs as an async job. Poll until it completes.
import time
while True:
status = requests.get(
f"https://api.empiriolabs.ai/v1/jobs/{job['job_id']}",
headers={"Authorization": "Bearer YOUR_EMPIRIOLABS_API_KEY"},
).json()
if status.get("status") in ("completed", "failed"):
print(status)
break
time.sleep(5)Request parameters supported by the SVI 2.0 Pro API on EmpirioLabs. Defaults apply when a field is omitted.
| 参数 | 类型 | 默认 | 范围 / 值 | 描述 |
|---|---|---|---|---|
| resolution | enum | 832x480 | 832x480, 480x832, 720x1280, 1280x720 | 480p is fast; 720p is slower but sharper. |
| duration | number | 18 | 18 to 121.5 | Estimated clip length in seconds. |
| cfg | number | 1 | 1 to 2 | Prompt adherence strength. |
| negative_prompt | string | vibrant tone, overexposed, static, blurry details, subtitles, style, artwork, painting, picture, still, overall gray, worst quality, low quality, JPEG compression residue, ugly, incomplete, extra fingers, poorly drawn hands, poorly drawn face, deformed, disfigured, malformed limbs, fused fingers, still picture, messy background, three legs, background crowd, walking backwards | - | Text describing what to avoid. |
| t2v_quality | enum | quality | fast, quality | Text-to-video pipeline tier. 'quality' uses the Wan 2.2 plus reference image model for higher fidelity; 'fast' uses the flash model for cheaper, quicker generations.... |
理论上是无限长的视频,带有一致的性格ID. 图像到视频通常得出优异结果 text-to-video. 约束 - 长视频的生成需要45+分钟 - 对于最佳动作:在提示格式 - jpg, jpeg, png, webp, heic, heif, bmp, tiff, tif 多场景模式** - 当在一个快速中描述多个场景时,下CFG(1.0-1.3)给予模型解释不同场景过渡的更多自由 - 提高CFG(1.5-2.0) 当每个场景必须按提示形式进行时,
On EmpirioLabs, SVI 2.0 Pro is billed pay as you go: 480p 视频 $0.057 每秒钟; 720p 视频 $0.17 每秒钟; T2V 快车 $0.065 额外费用. The live rate card on this page always matches what the API charges.
SVI 2.0 Pro is served through POST /v1/videos/generations on api.empiriolabs.ai with standard bearer-token authentication.
Yes. The EmpirioLabs playground runs SVI 2.0 Pro in the browser with the same parameters the API exposes, so you can test prompts before writing code.
Create an EmpirioLabs account, then generate a key under API Keys in the dashboard. Billing is pay-as-you-go credits, so you only pay for the requests you make.
Check out our pricing or reach out if you want your own model deployed on our stack.
