Google Drive or YouTube link as video_url: what Sume accepts
Some AI video tools accept Google Drive and YouTube links. Sume captions and face swap need a public HTTPS file URL. How to get your clip into that shape.
The short answer
Sume does not take a Google Drive or YouTube page link as video_url. Its Media inputs docs say inputs must be public HTTPS URLs, and they reject localhost, private-network, non-HTTPS, signed or private URLs and mismatched content types. HeyGen's lip sync tool (read 2026-10-04) says it accepts YouTube and Google Drive links alongside MP4, MOV and WEBM uploads up to 5 GB, so people arrive at Sume expecting the same.
The fix is to host the file at a plain public HTTPS address, then pass that address.
Why a share link is the wrong shape
A share link usually points at a viewer page, and a private share needs the signed-in viewer's permission. Both run against the rules above: a page that is not the video file is a content-type mismatch, and a link that needs a login is private. The Video captions docs repeat the constraint for POST /v1/video-captions: video_url must be a fetchable public HTTPS video URL. The face swap docs require the same of video_url for the Beta endpoint.
For YouTube links specifically, see the YouTube URL 422 post, which covers the unsupported-platform response.
What to do instead
Three routes get a clip to the right shape. Choose by where the file lives today.
- Already rendered on Sume: use its output URL. Sume mirrors outputs to
media.sume.com, which is public HTTPS by construction. - Stored in your own bucket or CDN: expose the object at a public HTTPS URL with a correct video content type, with no expiring signature in it.
- Only on a share service: download the file and re-host it, rather than passing the share page.
Side by side
The difference is what each product accepts as the source reference.
| Source | HeyGen lip sync tool | Sume captions and face swap |
|---|---|---|
| File upload (MP4, MOV, WEBM) | Yes, up to 5 GB | Pass a public HTTPS URL to the file |
| YouTube link | Accepted | Not accepted as video_url |
| Google Drive link | Accepted | Not accepted as video_url |
| Signed or private URL | Not stated on the page | Rejected |
Check it before you submit
Open the URL in a private browser window. If it plays the file without signing in, and the address ends at the video rather than a page, it is the right shape. Checking first saves a failed request. For face swap, also confirm the clip length, which the 4 to 15 second requirements post covers.
Sources
Related posts
More in Comparisons
- Ten holiday ad tracks: Lyria 3.5 at Google's list vs Sume Music
Google lists Lyria 3.5 at $0.08 a song and Sume Music bills $0.125 a generation, so ten holiday ad tracks cost $0.80 or $1.25. Here is what the gap covers.
- Image reference limits: 10 on ElevenLabs, 16 on Sume, 5 Ideogram 4.5
ElevenLabs lists 10 references for GPT Image 2.5. Sume's docs list 16 for the same models and 5 total for Ideogram 4.5. A table, a count guard and an edit rule.
- gpt-live-transcribe has no word timestamps: what it means for captions
OpenAI's realtime guide says gpt-live-transcribe cannot return word-level timestamps. If you need karaoke captions, that decides which path to use.
- Grok Imagine image: xAI $0.02 to $0.05 vs Runway 4 to 8 credits
xAI lists Grok Imagine images at $0.02, $0.04 and $0.05; Runway sells grok_imagine_image_2 at 4 to 8 credits ($0.04 to $0.08). Sume's router default is $0.025.
Written by Sume