3 steps to your first AI-finished video
Captions, dubbing, or translation — the path is the same. Follow along and see a finished file in about 10 minutes.
Upload footage
Open the studio and drag or pick a local video. MP4 / MOV / AVI / MKV. Aim for under 200MB; you can try 300–500MB. Free is capped at 15 minutes and rejects longer clips; Plus has no duration cap.
- No sign-up — upload with the anonymous account
- Detects speech, duration, and resolution
Pick one feature
Choose captions, dubbing, translation, or a cover, then upload. Each run does only that job. You can add more on the same video afterward.
- One feature per run — quota is charged for that tool
- Tweak results line by line online
Preview and export
When the preview looks right, download a captioned MP4 or an SRT file and publish to any platform.
- Hard-burn captions and publish as-is
- Export SRT / VTT caption files
AI caption tutorial:
timed captions in 10 minutes with a real timeline
The slowest part of manual captions is checking every timestamp. AI Video Factory hands that to AI — you upload and do the final proof. Full steps:
- 1
Upload a video and pick a caption source
In the studio pick “AI captions”, then upload. Transcribe speech (Mandarin / English), read on-screen text, or import an existing SRT / VTT.
- 2
Review the recognized lines
When recognition finishes, every line and timecode is listed. Skim quickly, click a wrong word, and edit it online. Check jargon, names, and brand terms.
- 3
Edit caption style visually
In the style panel, change font, size, color, stroke, and position. Every change previews live on the frame — what you see is what you get. No pro-software experience needed.
- 4
Export the video or caption file
When it looks right, export a hard-sub MP4 and publish, or export only SRT / VTT for another editor.
Works on phones too: open the studio in a mobile browser, pick a video from your camera roll, and follow the same steps. No app download.
AI dubbing tutorial:
turn copy into a polished voiceover
No mic, no booth. Turn copy or existing captions into natural speech and finish in minutes what used to take half a day.
- 1
Pick a dubbing source
Paste copy, or turn already-generated caption text into voiceover in one tap — no retyping.
- 2
Choose a voice
Chinese uses 晓晴. English can pick from several Kokoro voices. In auto-detect, Chinese lines use 晓晴 and English lines use the English voice you chose.
- 3
Adjust speed and Chinese parameters
Speed follows caption duration by default. You can also set speed by hand; Chinese dubbing can change tone and pace. Original audio can stay on.
- 4
Generate, mix, and export
After generate, preview the full mix. When you like it, mix with the original video on separate tracks and export.
Mix tip: if the video already has audio, keep the original track. Dubbing is written as an extra track and the picture stays locked to the source.
Caption translation tutorial:
30+ languages in one tap, go global easily
If a Chinese video needs overseas viewers, caption translation is step one. The flow continues right after captions are generated.
- 1
Generate or upload source captions
Follow the “AI caption tutorial” above for Chinese captions, or upload an existing SRT as the source.
- 2
Pick one target language
Choose the language in the dropdown — Chinese, English, Japanese, Korean, German, French, Russian, Spanish, Portuguese, and 30+ more. One target language per run.
- 3
Translate — keep the original timeline
Translation only changes the words. In and out points stay put. Spot-check jargon after it finishes.
- 4
Turn on bilingual captions if you need them
To show source and translation together, enable “Bilingual captions” in style settings. Two lines stack automatically.
- 5
Export this language version
Export a captioned MP4 or SRT. For another language, pick a new target and translate again.
Going-global tip: TikTok and YouTube tend to surface content fully covered in the local language. Bilingual captions usually earn more trust and watch-through than translation-only lines.
AI cover tutorial:
an eye-catching thumbnail in 3 minutes with smart frames
A good cover decides the click, but grab / retouch / typeset by hand is slow. AI Video Factory automates frame pick, layout, and multi-size export.
- 1
Upload or pick a video you already processed
Upload a new file, or pick one that already has captions or dubbing — no need to upload again.
- 2
AI picks keyframes
The system scores sharpness, faces, and motion across the cut and offers 6–8 highlight frames for you to choose.
- 3
Pick a template and export ratio
Choose a style — clean, big type, magazine, and more — and tick 16:9 / 1:1 / 9:16 as needed so one pass fits multiple platforms.
- 4
Add title text and a logo
Overlay a title, stickers, or your logo. Drag font, color, and position — live preview.
- 5
Export every size in one tap
When it looks right, export landscape, square, and vertical covers together. No redesign per platform.
Manual frame tip: if none of the AI picks work, scrub the timeline and grab any frame as the base. The system sharpens it, then you layout on top.
AI script / copy tutorial:
no footage yet? write the copy first
Not sure what to shoot? This entry does not need a video. Start from a keyword and AI turns the topic into a sectioned script you can use.
- 1
Enter a topic keyword
In the script tool, type a topic such as “fall outfits” or “workplace communication.” No writing background needed.
- 2
Pick a style and target duration
Choose talking-head, story, product, explainer, and similar templates, set a target length, and AI paces the copy to fit.
- 3
AI drafts a sectioned script
In a few seconds you get a full script with shot notes. Each section has an estimated duration so you can plan the shoot.
- 4
Edit the copy online
Rewrite a weak section, change tone, or add detail. Estimated duration recalculates after you edit.
- 5
Turn it into dubbing or a caption file
When the script is locked, convert it to AI voiceover or a timed caption file for the rest of production.
Best at the idea stage: this path is fully separate from “upload a video.” Keyword to finished copy never requires footage — ideal while you are still choosing a topic.
Private VPS deploy: Docker in one command
After you buy a VPS, follow these 4 steps — about 5 minutes — and you have a dedicated unlimited video-processing service.
Buy and log into the VPS
We recommend 2 cores / 4GB+ and Ubuntu 20.04+. Connect over SSH (Terminal / PuTTY).
Install Docker and Docker Compose
Most clouds offer a one-click Docker image. If it is not preinstalled, run the official install script — about 1–2 minutes.
Run the one-command deploy
Paste the command on the right into the terminal and press Enter. The script pulls images, sets env vars, and starts every service.
Open your private URL and go unlimited
When it finishes, the terminal prints a private URL. Open it on any device and use captions, dubbing, translation, and the rest without limits.
$ curl -fsSL https://get.example.com/deploy.sh | bash
[1/4] Checking system ... passed
[2/4] Pulling AI Video Factory image ... done
[3/4] Initializing config and volumes ... done
[4/4] Starting service containers ... done
✓ Deployed! Private URL:
https://your-vps-ip:8080If deploy fails: the usual causes are missing Docker or a busy port. Run docker -v to check the install. See “Troubleshooting” below for detailed fixes.
Problems people hit while following the tutorials
Bad error or a weak result? Try these first — most are fixable in a few minutes.
AI captions are inaccurate. What now?
Unsupported format or duration over the limit?
AI dubbing sounds unnatural or the speed is off?
Translated jargon is wrong. What now?
VPS deploy failed or Docker commands error out?
I finished the tutorial and still don't know what to click?
Follow the tutorial — ship in 10 minutes
Try every core tool free. Follow the steps and your first AI-processed video can go out today.
✓ Free quota on first visit · ✓ No card required · ✓ Runs in the browser