Video transcription

# transcribe.py
# Returns the transcript of a social video URL on any of seven platforms.
# Run with: INSIGHTSOCIAL_API_KEY=isk_live_... python transcribe.py
import os
from urllib.parse import urlparse

import requests

KEY = os.environ["INSIGHTSOCIAL_API_KEY"]
BASE = "https://api.insightsocial.app/v1"

ENDPOINT = {
    "youtube":   "youtube/video/transcript",
    "tiktok":    "tiktok/post/transcript",
    "instagram": "instagram/media/transcript",
    "facebook":  "facebook/post/transcript",
    "twitter":   "twitter/tweet/transcript",
    "reddit":    "reddit/post/transcript",
    "linkedin":  "linkedin/post/transcript",
}
HOSTS = {"youtube.com": "youtube", "youtu.be": "youtube", "tiktok.com": "tiktok",
         "instagram.com": "instagram", "facebook.com": "facebook", "fb.watch": "facebook",
         "x.com": "twitter", "twitter.com": "twitter", "reddit.com": "reddit",
         "v.redd.it": "reddit", "linkedin.com": "linkedin"}

def platform_of(url):
    host = urlparse(url).netloc.lower().removeprefix("www.").removeprefix("m.")
    for suffix, platform in HOSTS.items():
        if host == suffix or host.endswith("." + suffix):
            return platform
    raise ValueError(f"unsupported host: {host}")

def extract_text(platform, data):
    if platform == "youtube":
        return data.get("transcript_only_text") or ""
    if platform == "instagram":
        return " ".join(t.get("text", "") for t in data.get("transcripts") or [])
    t = data.get("transcript")
    if isinstance(t, str):
        return t
    # Other shapes: fall back to any plain-text field; inspect your first response.
    return data.get("transcript_only_text") or data.get("text") or ""

def transcribe(url):
    platform = platform_of(url)
    res = requests.get(f"{BASE}/{ENDPOINT[platform]}", params={"url": url},
                       headers={"x-api-key": KEY}, timeout=120)
    body = res.json()
    if not body["success"]:
        # A 404 here usually means "no captions" or "no speech", and costs nothing.
        return {"platform": platform, "text": None, "error": body["error"]["type"],
                "credits_used": body["credits_used"]}
    return {"platform": platform, "text": extract_text(platform, body["data"]),
            "credits_used": body["credits_used"]}

# Any public video URL on the seven platforms works; add your own to the list.
for url in [
    "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
]:
    t = transcribe(url)
    if t["text"] is None:
        print(f"[{t['platform']}] no transcript ({t['error']}), {t['credits_used']} credits")
    else:
        print(f"[{t['platform']}] {len(t['text'])} chars, {t['credits_used']} credits")
        print(t["text"][:200], "...")