Online and desktop
ClapClip online: 720p, 300 seconds, no account
What ClapClip online saves: the opening 300 seconds at up to 1280×720, in MP4, WebM or MP3, with no account, credits or media byte cap.
Key takeaways
The online downloader is deliberately narrow. Know the box before you paste the link, and you will not fight limits that are working as designed.
- Output is the opening 0–300 seconds at up to 1280×720, without upscaling, in MP4, WebM or MP3.
- MP4 and WebM require positive dimensions; MP3 reports 0×0 and needs a real audio track. No media byte ceiling applies.
- There is no login, credits or daily quota. Short-window rate limits, queue capacity and maintenance can still pause a task.
What does ClapClip online save, exactly?
ClapClip online saves the beginning of one public video or audio track: from 0 to at most 300 seconds, with video fitted into a 1280×720 box and never enlarged. Longer sources with a known duration are allowed, but only the opening 300 seconds are saved. That covers a reference clip, a quote or a sound check; it does not cover archiving a full episode.
You do not need an account, credits or a paid unlock. You do need permission to save the content.
Rule: if the part you need starts at minute 12, the online tool is the wrong tool. It always starts at 0 and stops at 300 seconds.
Which videos qualify?
Use one recorded, public, single-video link from a supported platform, with a known positive duration and permission to save. The supported list is YouTube (including Shorts), TikTok, Instagram (including Reels), X/Twitter and Reddit.
The online form is not for full-length recordings, custom start points, playlists, channels, profiles, private or login-only posts, live streams or DRM-protected media. It also cannot make a source accessible when that source refuses server requests.
Rule: one link, one video, one task, with a duration the source reports as positive, and only the first 300 seconds of output are saved. If you need a profile page or search result to find the video, open the individual post first.
What are the limits in numbers?
The numbers come from the site's download policy, not from marketing copy. The same checks run on the server that prepares the file.
- Maximum width 1280 px, maximum height 720 px. Smaller sources are not upscaled; a portrait video is at most 720 px tall.
- Maximum output duration 300 seconds (5 minutes). Video and audio both start at 0.
- No media byte ceiling. The source duration must be known and positive, and video formats must report usable dimensions.
- Three formats only: MP4, WebM and MP3. MP3 has 0×0 dimensions and requires a real audio stream.
The box is a ceiling, not a target. A 640×360 source stays 640×360 because the tool does not invent pixels, while a 1920×1080 source is fitted down to the box. A source with no fixed duration, such as a live stream, cannot satisfy the duration check at all.
Rule: an output wider than 1280 px, taller than 720 px, longer than 300 seconds, or with an unknown duration fails verification. A larger source is accepted only because the final output is constrained.
Why was my task rejected or stopped?
Rejections usually fall into four groups: the link, the source metadata, the output checks or capacity. Each has a different next step.
- Unknown, zero or negative duration: rejected before download, because the source did not report a usable length.
- Missing dimensions for MP4 or WebM: rejected before download, because the format did not expose usable width and height.
- Output outside 1280×720 or longer than 300 seconds: fails verification after preparation.
- Source refusal, expired analysis, no available worker, a full queue or maintenance: stops the task before or during preparation.
The rejection message is the useful part. An unknown-duration or invalid-dimension error will repeat until the source's metadata changes, while a capacity or maintenance error is usually temporary. Retrying a verification failure without changing the format changes nothing.
Rule: a failure that repeats after a fresh analysis is not a format problem. Change the link, wait for capacity, or accept that the source will not serve the request.
How do you get a file in five steps?
The workflow is short because the policy does most of the narrowing before you start. Every step below assumes you already have one public post you may save.
- Step 1 — Copy the public page or share URL of one recorded video you may save.
- Step 2 — Paste it into the form and run the analysis.
- Step 3 — Check the title, known duration and the formats actually returned.
- Step 4 — Choose MP4 or WebM for picture, or MP3 for sound, then start one preparation task.
- Step 5 — Save the ready file before the save link and the task record expire.
Pick the format by the dimensions you actually need, not by the biggest number in the list. If the analysis returns several MP4 options, all of them still have to fit the same 1280×720 box, and none of them adds detail the source never had.
Rule: do not open a second task for the same link while the first is running. Duplicate tasks do not move the queue faster, and the result will still be at most 300 seconds.
When is the desktop app the better choice?
Choose the online tool when the file you need is the opening 300 seconds at up to 1280×720 and you want no account, no installation and no byte cap. It handles one public video at a time and gives you MP4, WebM or MP3.
Choose the desktop app when you need a complete recording, a different start point or batch work. Rule: if the output you need runs past 300 seconds, start with the desktop workflow. It is a separate workflow, not an unlock for private or inaccessible content.
Bottom line: online is the shortest route for a short, permitted clip and the wrong tool for a full recording. The limits are the product: 300 seconds, 1280×720, three formats, no login and no byte ceiling.
References
Prepared with AI assistance using the listed official documentation and the website's current capabilities. Source platforms and tools can change.