Skip to main content

Documentation Index

Fetch the complete documentation index at: https://doc.hitopen.com/llms.txt

Use this file to discover all available pages before exploring further.

Video generation is an asynchronous operation. You submit a job with POST /v1/videos, receive a task ID, and then poll GET /v1/videos/{task_id} until the job is complete. Generation typically takes between a few seconds and several minutes depending on the model and video length.
Video generation is billed at job submission. If a job fails, contact your admin for a refund.

​
Create a video

POST https://YOUR_NEWAPI_BASE_URL/v1/videos
Submit a video generation job. The request uses multipart/form-data.

​
Request parameters

​
model
string
The video model ID. Supported models include Sora, Kling, and Jimeng variants. Contact your admin for the available model IDs on your instance.
​
prompt
string
A text description of the video you want to generate. The more detailed your description, the better the result.
​
image
string
An optional reference image provided as a URL or a base64-encoded string. When supplied, the model uses it as the starting frame or visual reference for the generated video.
​
duration
number
Desired video duration in seconds. The exact supported range depends on the model.
​
width
integer
Output video width in pixels.
​
height
integer
Output video height in pixels.
​
fps
integer
Frames per second for the output video.
​
n
integer
Number of videos to generate. Defaults to 1.
​
seed
integer
Random seed for reproducibility. Using the same seed and parameters tends to produce similar results.
​
response_format
string
The format of the response data.
​
metadata
object
Additional model-specific parameters passed as a JSON object. Common keys include negative_prompt, style, and quality_level.
​
user
string
An optional end-user identifier for monitoring and abuse detection.

​
Response fields (job object)

​
id
string
The task ID. Use this to poll for status with GET /v1/videos/{task_id}.
​
object
string
The type of the returned object.
​
model
string
The model that is processing the request.
​
status
string
Current job status. Typical values: "queued", "processing", "completed", "failed".
​
progress
number
A value between 0 and 1 indicating how far along the job is.
​
created_at
integer
Unix timestamp when the job was created.
​
completed_at
integer
Unix timestamp when the job finished. null if still in progress.
​
expires_at
integer
Unix timestamp after which the result URLs will no longer be accessible.
​
seconds
string
Duration of the generated video.
​
size
string
File size information for the generated video.
​
error
object
Present when status is "failed". Contains message and code fields.

​
Example

cURL
curl -X POST "https://YOUR_NEWAPI_BASE_URL/v1/videos" \
  -H "Authorization: Bearer sk-your-token" \
  -F model="sora" \
  -F prompt="A golden retriever running through a field of sunflowers at midday" \
  -F duration="5" \
  -F width="1280" \
  -F height="720"

​
Get video task status

GET https://YOUR_NEWAPI_BASE_URL/v1/videos/{task_id}
Retrieve the current status of a video generation job.

​
Path parameters

​
task_id
string
required
The task ID returned by the POST /v1/videos request.

​
Response fields

​
id
string
The task ID.
​
object
string
The type of the returned object.
​
model
string
The model processing the request.
​
status
string
Current job status: "queued", "processing", "completed", or "failed".
​
progress
number
Completion fraction between 0 and 1.
​
created_at
integer
Unix timestamp when the job was created.
​
seconds
string
Duration of the video (available once complete).

​
Example

cURL
curl "https://YOUR_NEWAPI_BASE_URL/v1/videos/task_abc123" \
  -H "Authorization: Bearer sk-your-token"

​
Async workflow

1

Submit the job

Call POST /v1/videos with your prompt and settings. Save the id from the response.
2

Poll for status

Call GET /v1/videos/{task_id} every few seconds. Check the status field.
3

Download the result

When status is "completed", the response will contain the video URL or data. Download it before expires_at.
Python (polling example)
import time
import requests

BASE_URL = "https://YOUR_NEWAPI_BASE_URL"
HEADERS = {"Authorization": "Bearer sk-your-token"}

# Step 1: Submit
job = requests.post(
    f"{BASE_URL}/v1/videos",
    headers=HEADERS,
    data={
        "model": "sora",
        "prompt": "A golden retriever running through a field of sunflowers",
        "duration": "5",
    },
).json()

task_id = job["id"]
print(f"Job submitted: {task_id}")

# Step 2: Poll
while True:
    result = requests.get(
        f"{BASE_URL}/v1/videos/{task_id}",
        headers=HEADERS,
    ).json()

    status = result["status"]
    progress = result.get("progress", 0)
    print(f"Status: {status} ({progress * 100:.0f}%)")

    if status == "completed":
        print("Video ready!")
        break
    elif status == "failed":
        print(f"Job failed: {result.get('error')}")
        break

    time.sleep(5)