It's shown most prominently in in Figure 4, which tracks post-weight-loss outcomes to 160 weeks (about three years). Sorry if it's not showing, I'm afraid sometimes I'm not sure what's open access anymore, since the journals have all done their best to make institutional logins automatic and invisible.

Practically all studies and reviews on the topic emphasize the large individual variability in outcomes. One in six Americans who has ever been overweight or obese loses at least 10 kilograms and keeps it off for at least a year, for example:

https://www.nature.com/articles/ijo201094

That study tracks a sample of adults in the general population, not a group assigned to diet. But evidently a significant fraction of people find success on their own. Most studies on the subject emphasizes substantial individual variability. Furthermore, a strong predictor of long-term weight-loss maintenance is just the initial weight lost:

https://onlinelibrary.wiley.com/doi/full/10.1111/ijcp.12300

https://academic.oup.com/ajcn/article/74/5/579/4737391

and another one is self-monitoring:

https://www.tandfonline.com/doi/full/10.2147/CIA.S25389

In another study, around 30% did not experience significant regain:

https://bpsmedicine.biomedcentral.com/articles/10.1186/s1303...

The notion that "all diets fail" seems to be perpetuated by a few unscrupulous journalists seeking clicks and social media slacktivists doing the same. It's not true, it's not helpful, and it's not defensible.