Reddit suspending aged accounts on manual iPhone farm - what changed this month?

saloed

Newbie
Joined
Apr 1, 2026
Messages
8
Reaction score
1
Running 3x iPhone 7+ (15.8.6, jb, Crane, 4G mobile proxies). Fully manual, no bots. Accounts used to last weeks/months fine, but lately they get suspended after some time - not hard banned, can still log in but account is flagged. Nothing changed on my end. Did Reddit update detection recently? Is Shadow tweak now essential? Anyone else seeing this pattern on older iOS setups?
 
A lot. Reddit algo is a mess nowadays. Fortunatelly, it still can be solved - not easy, but what safe road was built through easy ways?! https://www.blackhatworld.com/seo/my-reddit-account-creation-process.1689063/
Will read through it, thanks for the pointer
 
IF you're on mobile devices............ Your location/print is easier to identify.

Good luck.
 
Mobile devices can be a real pain when it comes to evading detection, especially with older iOS setups like iPhone 7+. It's likely that Reddit's algorithm has gotten better at fingerprinting mobile devices, which is why accounts are getting flagged even with manual operation and no bots. Using 4G mobile proxies helps, but it's not a foolproof solution. I've seen better results when combining mobile proxies with some extra layers of protection, like regularly switching user agents and mimicking real user behavior. It's also worth looking into the Shadow tweak, as it can help mask some of the device's identifiable features. From what I've tested, it's definitely made a difference in keeping accounts alive longer.
 
Mobile devices can be a real pain when it comes to evading detection, especially with older iOS setups like iPhone 7+. It's likely that Reddit's algorithm has gotten better at fingerprinting mobile devices, which is why accounts are getting flagged even with manual operation and no bots. Using 4G mobile proxies helps, but it's not a foolproof solution. I've seen better results when combining mobile proxies with some extra layers of protection, like regularly switching user agents and mimicking real user behavior. It's also worth looking into the Shadow tweak, as it can help mask some of the device's identifiable features. From what I've tested, it's definitely made a difference in keeping accounts alive longer.
Appreciate the info. Good to know Shadow tweak actually makes a difference in keeping accounts alive longer
 
Feels like Reddit tightened device fingerprinting and behavioral detection recently, especially on older jailbroken iPhone setups and reused proxy patterns. Common mistake is focusing only on bots while ignoring behavioral footprints and device signals that now seem heavily weighted.
 
Reddit is becoming harder and harder these days as there traffic is growing follow a few things dont post any thing in one day try to create a balance posting habit
 
When accounts stay login-accessible but flagged, that smells like trust throttling rather than a normal suspension
 
Sounds like Red*dit's cracking down. It's totally possible they've updated their detection system recently, and maybe Shadow tweak could help, but it's hard to say for sure without knowing all the details
 
While I'm running fully coherent with TLS proxy and mobile proxies, I'm still getting banned. My ML is still picking up the reverse-engineered bot signals, but what I've learned so far is that there's a lot of rigor on behavior of the account. How long they spend on reddit per day, session duration, interest graph, subreddit mix, etc. I'm getting ~5% of my freshly minted accounts through their first comments (abysmal rate) but that number is slowly climbing as the ML is fine-tuning the reverse engineered behavioral layer

You have to be at extremely high coherence, but ban rate is still ultra high at flawless coherence. Next frontier is behavioral/interest graph. I've picked up on some behavioral signals that I've proven reddit uses as "humanlike" - you're welcome to DM me, I don't want to dump my signals on the clear web, because I will then lose them.
 
Two thoughts: (1) what are you doing with the accounts - is there a pattern there? (2) is your hardware too unique? did JB give you a detectable, unique, fingerprint?
 
1. There is a rough pattern (interest area) of behavior; but it's pretty widely distributed in a broad space with relatively little overlap in interests. They are intended for advertising, but ease in with non-advertising comments. Even getting through the first non-advertising comment without getting hard banned is challenging.

2. Hardware isn't too unique; spoofed around common windows/chrome builds. No detectable/unique fingerprint; I've run pretty extreme correlation analysis on the potential fingerprints. I'm not getting linkage banned. Emails are unique and coherent

I've learned that reddit very intentionally lets your accounts get quite far before banning the account. They know that it's a bot far before the first comment, but wait until first comment to ban the account. Makes it time-consuming to reverse engineer. Bravo, reddit
 
Mobile devices can be a real pain when it comes to evading detection, especially with older iOS setups like iPhone 7+. It's likely that Reddit's algorithm has gotten better at fingerprinting mobile devices, which is why accounts are getting flagged even with manual operation and no bots. Using 4G mobile proxies helps, but it's not a foolproof solution. I've seen better results when combining mobile proxies with some extra layers of protection, like regularly switching user agents and mimicking real user behavior. It's also worth looking into the Shadow tweak, as it can help mask some of the device's identifiable features. From what I've tested, it's definitely made a difference in keeping accounts alive longer.
switching user agents on an aged account does more harm than good, reddit expects the same print every login so rotating it looks more sus not less. keep each account on one stable fingerprint and ip, the bans this month are pattern based not the device
 
Or just software spoof the native app - Android is a bit easier, I just got iOS spoofing working. It's a bit of a pain, but on-par with full-coherence web spoofing
 
1. use newer iPhones, iPhone 12 or higher.
2. JB and Crane can easily be detected. better to do 1 account per phone.
3. 4G proxies on iPhone I wouldn't recommend, better to use a real USA SIM card with data.
4G proxies ARE "real USA SIM cards" - they're egressing through verizon/tmobile/att IPs. There's no way to perceive the difference
Edit: Oh, you're selling 5G proxies by saying that 4G proxies are detectable as fake IPs lmao
 
4G proxies ARE "real USA SIM cards" - they're egressing through verizon/tmobile/att IPs. There's no way to perceive the difference
Edit: Oh, you're selling 5G proxies by saying that 4G proxies are detectable as fake IPs lmao
You clearly don't understand. Using a proxy on an iPhone with a proxy app is completely different than the iPhone using a physical SIM card.
You do realize some apps can detect that the phone dosent have a physical SIM card with data right? so just using a 4G/5G proxy may not be the end all solution.
Edit: No where I mentioned that 4G proxies are "fake IPs" lmao
 
Last edited:
Who is using proxies on hard phones? Are people actually running reddit at scale on phones? This seems massively unscalable

I'm currently work on iOS software (API) automation, and am close to full iOS coherence. Attest token signing is my biggest hurdle currently
 
Last edited:
4G proxies ARE "real USA SIM cards" - they're egressing through verizon/tmobile/att IPs. There's no way to perceive the difference
Edit: Oh, you're selling 5G proxies by saying that 4G proxies are detectable as fake IPs lmao
TikTok reads the SIM card number of the phone in order to determine the location. Obviously with a mobile proxy, this would not work and therefore they are not "real USA SIM cards". They're what they say they are: mobile proxies which proxy your data. Nothing more, nothing less.
 
Back
Top