Download videos from TLDV.io

tldv.py

from datetime import datetime
from os import system
import requests
import json

## Please install ffmpeg before running this script and make sure it's in your PATH
## brew install ffmpeg

## Please install requests before running this script
## pip3 install requests

## download video from tldv.io
##
## 1. Go to https://tldv.io/
## 2. Login
## 3. Go to the meeting you want to download
## 4. Copy the URL of the meeting
## 5. Open the developer tools (F12)
## 6. Go to the network tab
## 7. Refresh the page
## 8. Find the request to https://gw.tldv.io/v1/meetings/64145828ced74b0013d496ce/watch-page?noTranscript=true
## 9. Copy the auth token from the request headers
## 10. Run this script and paste the URL and auth token
## 11. python3 tldv.py

url = input("Please paste the URL of the meeting you want to download: ")

meeting_id = url.split("/")[-1]

print("\rFound meeting ID: ", meeting_id)

auth_token = input("Auth token: ")

data = requests.get(
    f"https://gw.tldv.io/v1/meetings/{meeting_id}/watch-page?noTranscript=true",
    headers={
        "Authorization": auth_token,
    },
)

try:
    response = json.loads(data.text)

    meeting = response.get("meeting", {})
    name = meeting.get("name", "No name")
    createdAt = meeting.get("createdAt", datetime.now())
    source = response.get("video", {}).get("source", None)

    date = datetime.strptime((createdAt), "%Y-%m-%dT%H:%M:%S.%fZ")
    normalised_date = date.strftime("%Y-%m-%d-%H-%M-%S")

    filename = f"{normalised_date}_{name}"
    filename_ext = ".mp4"

    command = f'ffmpeg -i {source} -c copy "{filename}.{filename_ext}"'

    json_filename = f'{filename}.json'

    with open(json_filename, "w") as f:
        f.write(data.text)

    print(command)

    print("Downloading video...")

    system(command)

except:
    print("Error encountered")
    print(data.text)

Anyone could download from MACOS? I have tried but it doesn't work. I am able to enter the url and the token but it seems to show me on terminal, all transcription of the video but does not download the video or anything.

Works like a charm!
I added a fork with parallel downloading in case you need to download and process multiple at a time.
I obtained links with selenium, and then made a list of videos. The MAX_WORKERS are basically the number of parallel videos to download and process. This makes the process way much faster.

Awesome works!

Steps:
Download fmpeg: https://www.youtube.com/watch?v=JR36oH35Fgg
Run the python script and it asks for url and token
Copy the tldv shared meeting link
Make sure the url at end doesn't have / after the meeting id
Take the authorization token as said above
Paste it and enter
ffmpeg starts downloadig... each bytes of video and saves as mp4 at the end

For people on Mac OS, I have added a Fork that works perfectly.
There was an issue in the name of the file created. It has an unsupported name due to '/'.

Anyone could download from MACOS? I have tried but it doesn't work. I am able to enter the url and the token but it seems to show me on terminal, all transcription of the video but does not download the video or anything.

Check my fork, it works on MacOS

Si desconozco todo lo que menciona porque no soy ingeniero de sistemas o relacionado. Deberia aplicar esto en mi laptop. me recomiendan algun tema en particular para aprender mas de esto por favor. se me van a borrar esos videos en 2 semanas y los necesito pero no puedo pagar el premium

import os
import re
import json
import requests
from datetime import datetime
from subprocess import run, CalledProcessError
from tqdm import tqdm
#open cmd and type pip3 install requests tqdm

--- Helper Functions ---

def sanitize_filename(name):
return re.sub(r'[\/*?:"<>|]', "_", name)

def get_meeting_id(url):
return url.rstrip("/").split("/")[-1]

def fetch_meeting_data(meeting_id, auth_token):
api_url = f"https://gw.tldv.io/v1/meetings/{meeting_id}/watch-page?noTranscript=true"
response = requests.get(api_url, headers={"Authorization": auth_token})
response.raise_for_status()
return response.json()

def download_video(source_url, output_file):
command = [
"ffmpeg",
"-i", source_url,
"-c", "copy",
output_file
]
try:
run(command, check=True)
print(f"\n✅ Download completed: {output_file}")
except CalledProcessError as e:
print("\n❌ Error during download.")
print(e)

--- Main Logic ---

def main():
print("📥 TLDV Video Downloader")

url = input("Paste the full TLDV meeting URL: ").strip()
if not url:
    print("⚠️  URL cannot be empty.")
    return

auth_token = input("Paste your auth token: ").strip()
if not auth_token:
    print("⚠️  Auth token is required.")
    return

meeting_id = get_meeting_id(url)
print(f"🔍 Found Meeting ID: {meeting_id}")

try:
    data = fetch_meeting_data(meeting_id, auth_token)
except requests.exceptions.HTTPError as err:
    print(f"❌ Failed to fetch meeting data: {err}")
    return

meeting = data.get("meeting", {})
name = sanitize_filename(meeting.get("name", "TLDV_Meeting"))
created_at = meeting.get("createdAt", datetime.now().isoformat())

try:
    date_obj = datetime.strptime(created_at, "%Y-%m-%dT%H:%M:%S.%fZ")
except ValueError:
    date_obj = datetime.now()

timestamp = date_obj.strftime("%Y-%m-%d-%H-%M-%S")
output_filename = f"{timestamp}_{name}.mp4"

source_url = data.get("video", {}).get("source")
if not source_url:
    print("❌ Video source not found.")
    return

print(f"\n🎬 Title: {name}")
print(f"📅 Created At: {timestamp}")
print(f"🎯 Output File: {output_filename}")

confirm = input("\nProceed with download? (y/n): ").strip().lower()
if confirm != 'y':
    print("❌ Download canceled.")
    return

json_backup = f"{timestamp}_{name}.json"
with open(json_backup, "w") as f:
    json.dump(data, f, indent=2)
    print(f"📝 Metadata saved to {json_backup}")

print("⏬ Downloading video...")
download_video(source_url, output_filename)

--- Run ---

if name == "main":
main()

Works like a charm! I added a fork with parallel downloading in case you need to download and process multiple at a time. I obtained links with selenium, and then made a list of videos. The MAX_WORKERS are basically the number of parallel videos to download and process. This makes the process way much faster.

I've used this from javier-ortizt
I'm a newbie here, but I'm so glad I got it to work haha! Basically, I installed ffmpeg on my Mac using Homebrew (I followed this tutorial) and once it's set up, I ran the command

python3 tldv.py

Thanks so much!!