AI Tools
AI Video Script Generator
Get a full spoken script for your video topic — hook, intro, main points with B-roll notes, recap and call to action.
Open toolUrdu Caption Generator
Most Urdu and Hinglish videos on YouTube, TikTok and Reels ship without subtitles — which quietly shuts out everyone watching on mute, everyone hard of hearing, and every viewer whose Urdu is stronger in one script than the other. This generator fixes that in one pass: upload a video or audio clip (or paste a YouTube link, or just the spoken words), and the AI transcribes the speech into timed caption lines, each one written twice — in proper Urdu script and in Roman Urdu, the way Urdu is actually typed on phones across Pakistan. Download the Urdu SRT, the Roman Urdu SRT or a VTT file and drop it straight into YouTube Studio, CapCut or Premiere. Mixed Hinglish speech is handled the way it is spoken: Urdu words stay Urdu, English words stay English in the Roman line. Like every transcription, it deserves one honest proofread — names, numbers and places are where any AI can slip — but the heavy work of listening, timing and writing both scripts is done for you.
MP3, MP4, WAV, M4A and most phone recordings work. Best results under about 12 MB — a few minutes of speech. The file is read in your browser and sent to the AI only to create your captions; it is not stored.
Link transcription is best-effort on the free tier. If a link fails, download the video and use Option 1 instead — uploads are the reliable route.
Use one option only. If you fill in more than one, the file wins, then the link, then the text. Pasted text gets estimated timings (there is no audio to measure); uploaded media gets real timings.
This tool is powered by AI running on a free service tier, so a generation can occasionally take a little longer or ask you to retry at busy times. Your input is sent only to produce your result — it is not stored.
Honest limits: this runs on a free AI tier, so very long or noisy recordings can fail or need a retry. For a long video, trim it into parts with the Video Trimmer, or extract the audio first with Video to MP3 — smaller files transcribe faster and more reliably.
How to use this tool
- Choose your input: upload an audio or video file, paste a YouTube link, or paste a spoken transcript.
- Pick the spoken language — Auto-detect, Urdu, or Hinglish (Urdu mixed with English).
- Click Generate captions and wait while the AI transcribes — longer clips take longer.
- Check the caption table, then download the Urdu SRT, the Roman Urdu SRT or the VTT file and add it to your video.
Worked example
A creator uploads a three-minute Urdu vlog clip recorded on a phone. The tool returns a table of timed caption lines — each line written once in Urdu script and once in Roman Urdu — plus ready SRT and VTT files that drop straight into YouTube Studio, CapCut or Premiere as the video's subtitles.
Frequently asked questions
Urdu script or Roman Urdu — which subtitle file should I upload?
Upload both if your platform allows more than one subtitle track (YouTube does): the Urdu-script track serves fluent readers, the Roman Urdu track serves the huge audience that speaks Urdu but reads Latin letters all day. If you can only pick one, pick the script your own audience types in — for most Pakistani social audiences that is Roman Urdu; for news, education and family audiences it is often Urdu script.
How well does it handle Hinglish — Urdu mixed with English?
That mix is the normal case this tool is built for, not an edge case. In the Roman Urdu line, English words keep their normal English spelling ('meeting', 'deadline') because that is how people type them; in the Urdu-script line the same words are written the way they sound. Pick 'Hinglish' in the language menu when your speech mixes heavily — it noticeably improves the split.
What file size and length work best?
On the free AI tier, clips under about 12 MB — typically a few minutes of audio, or a short compressed video — are the sweet spot. For a long video, cut it into parts with the Video Trimmer and caption each part, or extract the soundtrack with Video to MP3 first: an MP3 of a 20-minute talk is often smaller than a 2-minute phone video, and audio is all the transcription needs.
How do I add the SRT file to my video?
On YouTube: open the video in YouTube Studio, choose Subtitles, pick the language, then 'Upload file' and select the SRT. In CapCut and Premiere, import the SRT as a caption/subtitle track and it lands on the timeline already timed. The VTT file is the format web players (and some editors) prefer — same captions, different envelope.
Is my recording stored or kept?
No. Your file is read in your browser, sent once to the AI service to produce these captions, and is not stored by this site. The same caution applies as with any AI service, though: do not transcribe recordings that are private or sensitive enough that you would not want them processed by a third-party AI at all.
Why do pasted transcripts get 'estimated' timings?
A transcript has no clock — nothing in plain text says when each line was spoken. So for pasted text the tool estimates timings at a natural speaking pace to give your editor a working starting track. Upload the actual audio or video whenever you can: then the timings are measured from the media itself and land much closer.