Upscale an old Sora download: the limits of Sume's video upscaler
Sora files you saved can be upscaled. Sume's video-upscale model takes a scale from 1.1 to 4, up to 30 seconds, and a fast, standard or pro tier.
If you downloaded Sora clips before the Videos API shut down on 2026-09-24 (OpenAI deprecations page, read 2026-10-03), you can still improve them. Sume's video upscale endpoint, POST /v1/video-upscale-1.0/upscale with model sume/video-upscale-1.0, takes a video_url and a scale and returns a larger clip. Upscaling adds pixels, it does not recover detail the render never had.
Request fields
Set duration_seconds to the clip's real length. The default reserves 5 seconds, which is wrong for most clips. The same operation is available to agents as the video_upscale_create tool. Pricing is shown in your workspace, not repeated here.
| Field | Value |
|---|---|
| video_url | Source clip URL |
| upscale_factor (or scale_ratio) | 1.1 to 4 |
| duration_seconds | 1 to 30; default 5 is reserved |
| enhancement_tier | fast, standard or pro |
Sensible settings
The existing upscale guides cover 4K targets and tier choice in more depth.
- Start with a 2x factor and the standard tier on one clip before running a folder.
- Keep the original file; upscaling a re-encode of an upscale compounds artifacts.
- If the clip is longer than 30 seconds, cut it first with video-trim.
Sources
Related posts
More in Media tools
- Vertical video subtitles: BBC's 3-line rule and Sume settings
BBC guidance for 9:16 subtitles: up to 3 lines, 90% width, placed a little high. How to set safe_width_ratio and anchor_ratio on Sume burned-in captions.
- Vimeo "Invalid Caption File": build clean WebVTT from Sume
Vimeo rejects a cue that starts at the previous cue's end. Sume sentence segments touch, so trim 1 ms, write UTF-8 WebVTT and upload. Script included.
- Webinar to Shorts: pick moments by transcript, not fixed cuts
Fixed-interval cuts make interchangeable Shorts. Read the webinar's sentence segments, choose complete thoughts, then trim and caption each with Sume.
- WebVTT cue settings (line, size) vs Sume anchor_ratio and width
WebVTT line:78%,center and size:90% have close matches in Sume's caption design fields. A converter script, and what per-cue settings a burned render loses.
Written by Sume