Skip to main content
POST
/
serve
/
api
/
v1
/
download
Download Video
curl --request POST \
  --url https://api.memories.ai/serve/api/v1/download \
  --header 'Authorization: <api-key>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "video_no": "VI625239098370850816",
  "unique_id": "default"
}
'
"<string>"

Documentation Index

Fetch the complete documentation index at: https://api-tools.memories.ai/llms.txt

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

Product: Visual Search Use case: Upload videos and images, auto-index them, then search by natural language, image, or transcript phrase Host: https://api.memories.ai/serve/api/v1 Auth: Authorization: sk-mavi-... (no Bearer prefix)
Download the original video file for a given video in your Private Video Library. Returns a binary stream on success.

​
Prerequisites

​
Endpoint

POST /serve/api/v1/download

​
Authentication

Include your API key in the Authorization header. 
Authorization: sk-mavi-...

​
Request

Content-Type: application/json 
{
  "video_no": "VI625239098370850816",
  "unique_id": "default"
}

​
Request Parameters

​
video_no
string
required
The video identifier to download.
​
unique_id
string
default:"default"
Scope identifier for the authenticated account. Must match the unique_id used at upload time.

​
Python Example

import requests

headers = {"Authorization": "sk-mavi-..."}
payload = { 
    "unique_id": "default", 
    "video_no": "VI625239098370850816"
}
url = "https://api.memories.ai/serve/api/v1/download"

# stream=True enables chunked download
resp = requests.post(url, json=payload, headers=headers, stream=True)
print("Status:", resp.status_code)

content_type = resp.headers.get("Content-Type", "")

if "application/json" in content_type.lower(): 
    # JSON means the download did not succeed; inspect the error payload 
    print(resp.json())
else: 
    # Binary stream (video) 
    with open("video.mp4", "wb") as f: 
        for chunk in resp.iter_content(chunk_size=8192): 
            if chunk: 
                f.write(chunk) 
    print("Download complete")

​
cURL Example

curl -X POST "https://api.memories.ai/serve/api/v1/download" \
  -H "Authorization: sk-mavi-..." \
  -H "Content-Type: application/json" \
  -d '{"video_no":"VI625239098370850816","unique_id":"default"}' \
  --output video.mp4

​
Successful Response

On success, the server returns the video file as a binary stream:
  • Status: 200
  • Content-Type: application/octet-stream
  • Body: Binary content of the video (mp4 — the first bytes are an ftyp ISO-BMFF box)
The download uses HTTP chunked transfer encoding, so Content-Length is typically absent. Don’t use it for progress reporting; rely on iter_content chunk counts instead.

​
Error Response (JSON)

If the download cannot be fulfilled, a JSON payload is returned instead — same HTTP status (200) but Content-Type: application/json. The Python and cURL examples above already branch on the response content type. Unknown or wrong-namespace video_no: 
{
    "code": "0001",
    "msg": "download video error",
    "data": null,
    "success": false,
    "failed": true
}
Missing video_no in the request body: 
{
    "code": "0001",
    "msg": "The video number cannot be empty！",
    "data": null,
    "success": false,
    "failed": true
}
Branch on Content-Type (or check success: false) rather than parsing msg text — the message strings are short, sometimes use full-width punctuation, and aren’t a stable contract.

​
Tips

  • Stream downloads: Always use stream=True (Python) to avoid loading the entire file into memory.
  • Scopes: If you organize content by unique_id, include it for correct scoping.

Authorizations

​
Authorization
string
header
required

Body

application/json
​
video_no
string
required
Example:

"VI625239098370850816"

​
unique_id
string
default:default

Response

Binary video stream or JSON error response

The response is of type file.