How to Extract Hardcoded Subtitles from Video

Why this matters

Hardcoded subtitles are pixels, not text — extraction is an OCR problem with a temporal twist: the same line persists across many frames and must be emitted once.

Per-frame OCR alone produces thousands of duplicate lines; the pipeline clusters them by content and timing to produce clean, readable SRT.

How it works

Submit the video, choose output_format (srt/vtt/ass), and poll. The API returns the subtitle file plus per-line confidence scores.

Submit a job with an input URL (S3, GCS or HTTPS), poll the job URL, and download the rendered output. No GPU, no queues, no ffmpeg builds to babysit.

Code

import requests, time

API = "https://api.mlslabs.io/v1/subtitle-extractor/jobs"
headers = {"X-API-Key": "YOUR_API_KEY"}
payload = {"input": "s3://bucket/video.mp4", "output_format": "srt"}

resp = requests.post(API, json=payload, headers=headers)
job = resp.json()
while job["status"] not in ("succeeded", "failed"):
    time.sleep(3)
    job = requests.get(job["url"], headers=headers).json()

print("Output:", job["output_url"])

Pro tips

• Higher source resolution means better OCR — upscale first if the source is low-res.
• Set the source language when you know it; it improves character recognition.
• Review low-confidence lines — usually the ones with partial occlusion or stylized fonts.

Try the extraction flow in Postman.