Idk what's your case and maybe it will get better by itself eventually, but if your phone REALLY got marked in their system, then reinstalling the app won't help.
Phones have IDs - on IOS you can't change it (I think), on Android you can, but as far as I remember you need root for that, and root...
did you register accounts those phones before, did they get banned, do you use proxies, how do you warmup the accounts, are the reels unique and were never posted before?
Ask Claude to create a script to add (almost) invisible visual noise, resize the picture a bit and clean metadata at the same time. The noise is the key.
Real browser or phone automation is the only safe way, all platforms can surely detect raw requests if they want. It does eat traffic, but you can block media or switch to ISP proxies for like 4$/pc. What do you need the automation for?
I'd bet on owning distribution and trust - an audience, a client base, partnerships, things where someone has to be accountable. But who knows what will actually happen, the world is crazy rn...
Hey all,
I built a fully autonomous bot farm for X, Reddit, Instagram and Threads. Accounts get created, warmed up and post on their own (phone+browser hybrid automation, different campaigns supported, day windows, full humanization etc.), I barely touch it. Until recently I was using it to...
I've seen people using gemini or deepseek vision on a screenshot with overlayed numbered grid so it can give more exact results for recaptcha/hcaptcha before.
Now the problem isn't really these captchas but WAFs.
I mean they still give you a cheap access to antidetect phones even tho they're virtual. The next level is to get physical phones, but it's much more expensive and not always the case when you need the quantity.
Gmail + phone.
Works, but I buy gmail for 1.5-2$/pc and not all of them are valid so it's apparently not the most efficient solution.
About 30% of my gmail accounts also ask for captcha just as you said and I automated my bots to solve it - sometimes works and sometimes it says smth like "try...
I have an Instagram/Threads farm on Geelark. It's like antidetect browser but phone, they don't get detected if you setup them right but a bit expensive to run.
surely not, it does zero for ban evasion. reddit links accounts by fingerprint + IP + behavior.
do you register the accounts by yourself or buy registered? when exactly the bans happen? what proxies do you use? there are lots of things to dig into.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.