Captions only
Needs no key or model
Clipverb takes the captions the video already has, breaks them into paragraphs and starts each paragraph with its time in the video. Each time links to that moment. You only need to pick a folder.
Clipverb files the video's own captions and links each paragraph to its moment in the video. Or it sends the captions to the AI provider you pick, on your own key, and files the summary.
Clipverb is in development, and you can't download it yet.
---title: How memory workssource: youtube.com/watch?v=...tool: Clipverbmode: captions---
# How memory works
[0:03]Memory starts with attention: your brain stores what you notice.
[1:42]Sleep moves the day's traces into long term storage.
[6:15]Testing yourself beats reading the same page again.
How it works
Needs no key or model
Clipverb takes the captions the video already has, breaks them into paragraphs and starts each paragraph with its time in the video. Each time links to that moment. You only need to pick a folder.
On your own key
Clipverb sends the captions to the AI provider you choose, on your own API key, and files the summary it gets back. The summary keeps the how and the why, the key claims with their timestamps, the numbers worth keeping and the parts worth watching again.
Clipverb writes ordinary Markdown with YAML frontmatter into the folder you already use, such as an Obsidian vault. If a file with the same name exists, Clipverb gives the new note a new name.
Clipverb saves its queue to disk, so your links survive a restart. When a link fails, Clipverb tries again after a longer wait each time. If it still fails, Clipverb shows the error on that link's row.
Pick a model that runs on your computer through Ollama, and the transcript stays on your computer.
Your notes and your key
Clipverb reads the video's caption track and leaves the video and its audio alone. If a video has no captions, Clipverb can't file it in either mode.
Clipverb asks the video site for the caption track. In Captions + distill, it also sends the captions to the AI provider you picked, on your key. Clipverb has no account, sign in, telemetry, crash reporter or update check.
The desktop app stores your API key in your computer's credential store. Your provider bills you directly for what you send.
Clipverb saves each note as an ordinary file in the folder you chose, so any app can open it.
Pricing
Clipverb charges once, with no subscription and no fee per video. Captions only needs no key. For Captions + distill, you use your own API key or a local model, and your provider bills you for what you send it.
Questions
Captions only costs nothing more: you need a folder and nothing else. In Captions + distill, your AI provider bills you directly for what you send it.
Clipverb saves them in a folder you pick, as ordinary Markdown files with YAML frontmatter.
Clipverb has no servers. In Captions only, the transcript stays on your computer. In Captions + distill, Clipverb sends it to the AI provider you chose, on your own account. The privacy page lists each place Clipverb connects to.
Clipverb works with videos that have captions, in either mode. It files the captions the video already has, so it can't file a video with no caption track.
We are building a web version now.