YouTube Description Generator From Your Video
BetaA YouTube description generator that starts from what you actually say in the video. Drop the file and ProMedia transcribes it on your device, then drafts a title, a description and hashtags, or a TikTok, Instagram or X caption, in the tone you pick. Edit the draft and post. It is in Beta, and nothing is uploaded.
- Runs on your device
- Whisper AI
- Works offline
Updated
This tool is in beta: it works, but it is heavy for some devices and results may need a second look.
Drop your videos or audio files here
How to write a YouTube description from your video
Step 1: Drop your video, or its audio, on the box above.
Step 2: Pick the Platform (YouTube, TikTok, Instagram or X), a Tone and how many Hashtags you want (0–15).
Step 3: Press Create and confirm the one-time speech model download. The video is transcribed, then the draft is written.
Step 4: Read the draft, change anything that is not quite you, and copy it into the upload form.
Good to know
Settings explained
- Spoken language
- Picking the language skips detection and improves accuracy on short clips.
Formats
Batch-friendly: add many files and they are processed one after another — if one fails, the rest still finish.
How it runs
- Whisper speech model
- Speech recognition uses an open Whisper model. It downloads once after you confirm its size, then runs on your device — on the GPU via WebGPU when available.
What you get for each platform
YouTube gets a separate title and description, so the tool doubles as a YouTube title generator; TikTok, Instagram and X get a single caption with a hook line first. Every draft is trimmed to fit the platform, with the hashtags at the end, and saved as a text file you can keep with the video.
| Platform | Draft | Kept under |
|---|---|---|
| YouTube | Title, description, hashtags | 100 characters (title), 5,000 (description) |
| TikTok | Hook line, caption, hashtags | 2,200 characters |
| Hook line, caption, hashtags | 2,200 characters | |
| X | One short post with hashtags | 280 characters (standard accounts) |
How the draft is written
Everything starts with a transcript made by Whisper in your browser. What happens next depends on your browser. Where desktop Chrome lets the page use its built-in language model (Gemini Nano), that model writes a fresh title and description from the transcript and is instructed not to invent facts that are not in it. Other browsers get a keyword draft: the most informative sentences of the transcript, its key phrases as hashtags and a closing line that matches the tone.
| Chrome on-device AI | Keyword draft | |
|---|---|---|
| Where it runs | Inside desktop Chrome, on your computer | In any browser, on your device |
| Title | Written for the platform, under 70 characters for YouTube | Built from the two strongest key phrases |
| Description | 2–4 sentences for YouTube, a short caption elsewhere, in your tone | The most informative sentences, plus a closing line |
| Hashtags | Suggested by the model | Key phrases turned into hashtags |
How to write a good YouTube description
A good description tells a viewer, and YouTube search, what the video covers in the words people would type. Use the draft as the first two lines, then add the rest yourself.
- Lead with the topic. Only the first lines show before viewers expand the description, so say what the video is about there.
- Use the words your audience searches. The draft takes them from what you say, which keeps them honest.
- Add chapters. A timestamp list from the YouTube chapters generator helps viewers and can appear in Google results.
- Link what you mention. Products, sources and earlier videos belong below the summary.
- Finish with a few hashtags. YouTube shows up to three of them above the title.
Hashtag generator for video: why hashtags matter
Hashtags group a post with others on the same topic and make it findable by people who search or tap that tag. More is not better: YouTube ignores every hashtag on a video that has more than 60, and three to five relevant tags read better than a wall of them on any platform. The slider goes up to 15 and defaults to 5.
This tool is not a type-a-topic hashtag generator: it needs your video, because its hashtags come from what is said in it.
What the Beta can and cannot do
- It writes from the transcript only. It cannot see the picture, so a silent video has nothing to write about.
- Quality depends on the browser. Desktop Chrome with on-device AI writes the most natural copy; elsewhere the keyword draft is a starting point.
- It drafts in the spoken language when Chrome’s model supports it; the keyword draft always uses the transcript’s own words.
- Private. No video, transcript or draft is sent to an AI service.
Tips & troubleshooting
Captions for the video itself are a different job: the subtitle generator puts the spoken words on screen.
Cutting a long video into Shorts first? Make the clips with the AI clip maker, then write a caption for each.
Want the whole transcript as a starting point for a blog post? Use video to text.
Run it twice with different tones and combine the best lines of each.
Post Text (Caption & Hashtag Writer) FAQ
How do I write a good YouTube description?
Say what the video covers in the first two lines, using the words viewers would search for, then add chapters, links to anything you mention, and up to three hashtags. This generator drafts the opening lines and hashtags from your own transcript, so you start from accurate wording.
What should I put in my YouTube description?
A short summary of the video, chapter timestamps for anything longer than a few minutes, links to products, sources or related videos, and a few relevant hashtags at the end. Keep the most important sentence first, because the rest is hidden until viewers expand it.
How do I write a YouTube description for SEO?
Describe the video plainly with the terms people search, near the start, and do not stuff keywords. A transcript-based draft helps because it uses the words you actually say. Chapters add searchable section titles, and YouTube’s own title and description limits keep you concise.
Is my video sent to an AI service?
No. The transcript is made by a speech model running in your browser, and the draft is written either by Chrome’s built-in on-device model or by a keyword method in the page itself. Nothing is sent to a cloud AI.
Why are hashtags important?
They connect your post to a topic, so people who search or follow that hashtag can find it. On YouTube, up to three appear above the title. Use a few specific tags rather than many generic ones; YouTube ignores all hashtags on a video that has more than 60.
Why is Post Text in Beta?
Its quality depends on your browser. Desktop Chrome’s on-device language model writes natural copy; other browsers get a simpler draft built from key phrases and sentences of the transcript. Treat every draft as a first version and edit it.