I was using a chrome extension to bulk download my Sora 2 drafts. I checked the code on the extension’s GitHub and noticed that it programmatically creates links for unshared drafts and so now when I try to post or create a shareable link to a draft video, I get a “Failed to create link” or “Too many requests” error. Anybody know how long the rate limit lasts? It’s been 10 hours since I noticed this and I’m still unable to post and create shareable links to my drafts so they can be downloaded.
Hey @dwight_bauer, yeah, that’s frustrating, especially when it blocks something simple like sharing drafts.
This usually happens when a burst of link requests (like from that extension) pushes you into stricter rate limits. They use rolling windows, so repeated spikes can extend the cooldown, sometimes closer to ~24h.
Since you’re ~10 hours in, you might still be in that window. Best move is to keep the extension off, wait it out a bit longer, and retry from a clean session.
To avoid it later, space out requests and avoid auto-retries.
More context here: Rate limits
If it’s still blocked after a full day, it’s likely something else.
Curious if it clears up soon.
-Mark G.
So you were right, the rate limit timeout did end after 24 hours. But now what’s bullshit is this time I just created links for my drafts manually one by one which is definitely a slow pace, yet even just doing this after a while still leads you to getting limited. So now I’m waiting in limbo again wondering if this will also be a 24 hour timeout, or if it’ll be a shorter one since this wasn’t a burst request rate like last time smh ![]()
Hi @dwight_bauer, yeah, that does feel rough especially if you already slowed things down and still got limited again.
What’s likely happening is limits aren’t just about speed. Rate limits are measured in five ways: RPM (requests per minute), RPD (requests per day), TPM (tokens per minute), TPD (tokens per day), and IPM (images per minute). Rate limits can be hit across any of the options depending on what occurs first.
Also, these limits apply at the org or project level, not per user, so other activity can count toward it without being obvious. And the timeout isn’t always 24 hours, it depends which limit you hit.
If you can, check your limits page and consider adding exponential backoff so retries don’t stack into the limit. If it keeps happening, worth flagging so they can confirm exactly which cap you’re hitting.
-Mark G.