[Attention] To proxy sellers and instagram marketers

[thetrustedzone]

Something I got multiple PMs about and also noticing in BHW posts ... Instagram accounts banning and Instagram proxies ...etc

here is my 2 cents :

Instagram has IPs database check against ... usually, they know if you are using proxy or not but not always.
when they detect you using a proxy, they don't block you like other platforms but instead, they are watching you carefully and apply more anti-bot factors against your account and this is one of the main reasons why automated IG accounts get banned quickly nowadays ... to understand that simply imagine you are refugee entered US country with a fake passport and for a reason FBI knew you used a fake passport to enter and allowed that BUT for your first mistake in the US you should expect FBI through you outside the US while if entered with a legit passport that mistake probably will not be noticeable! the passport here for the FBI same as your IP address for Instagram.

So how to know if your proxy is detected by Instagram or not?


We could do different tests to know that, simple and easy one -which also answer why some proxies do not work for scraping-, is sending an HTTP request with the proxy against Instagram public API end and watching the result: if Instagram redirects the request to login page this means the proxy is detected if Instagram didn't redirect and retrieved JSON data this means the proxy isn't detected.

Most Proxy providers when to assure you they selling working and good Instagram Proxies ... they usually test like this: send an HTTP request with the proxy to Instagram host if response status code 200 ok and timeout passed => good Instagram proxy, right? WRONG

If we want to test correctly the proxy is good and not detected by Instagram technically we do like this:

send HTTP request to IG API end > if got JSON response it's good and not detected, else it's detected proxy and not good for IG.

code example translates above(in c# but could do the same in any other programming language) :

 var httpClientHandler = new HttpClientHandler()
                {
                    Proxy = new WebProxy("127.0.0.1:8080", false),
                    UseProxy = true
                };

                var client = new HttpClient(handler: httpClientHandler, disposeHandler: true);
                var response = await client.GetAsync("https://www.instagram.com/instagram/?__a=1");

            
                if ( response.Content.Headers.ContentType.MediaType == "application/json")
                {
                    // Not detected proxy and good for Instagram

                }
                else
                {
                    Detected proxy and not good for Instagram
                }

Don't worry I know most of you not coders that code above provided for serious proxy providers want to improve the service for their clients and for junior bots coders... I made a simple online tool to test that here, it's part of my beta test here and in the next update the client desktop client will contain a similar IG proxy checker.. feel free to use any other tool test the same concept.


So now which proxy providers and which proxy type is good and not detected by Instagram?

in my tests, I found most of the data center proxies detected and not good for IG, while most of the rotating residential proxies including 4 g mobile proxies are good and not detected by Instagram.

@Saly Verona wrote an excellent article about the proxies type and why rotating residential including 4g mobile is good for social media.
I highly recommend reading it particularly if you are using automated tools.

BHW Proxy Marketplace full of sellers offers free trial test them before purchase.


If you have any questions/comments please reply in this thread and don't pm me it.

 

[El Duque]

Thanks for the advice sir

 

[Standartuye]

Great post!

Does this proxy rule apply to scrape with accounts?

Which one would be faster for scraping? Using lots of accounts with proxies or using only proxies? What one do you plan to implement on your software?

 

[Deleted member 1107112]

Thanks for the share

 

[Christion]

Nice share! Thanks mate

 

[GrowthStrategist]

Nice share. That said I'm doing fine scraping with ipv6 datacentre. To a certain degree anyway... Just not sure how much worth it buying 500 residential IP's is against 500 datacentre ipv6 (quite a lot more money)

 

[DunderMannen]

Would make sense. Watching accounts you know are bots makes your AI even more powerful in the future!

 

[thetrustedzone]

Great post!

Does this proxy rule apply to scrape with accounts?

Which one would be faster for scraping? Using lots of accounts with proxies or using only proxies? What one do you plan to implement on your software?

1. Yes this applies to scrape with accounts if proxy detected don't use it with/without accounts for Instagram just pick another proxy type/provider, using accounts with detected proxies guaranteed way to get high accounts banning rate.

2.Regarding Instagram scraping there's mainly two types of data :

A.Data could extract without login: like user basic info NO. of followers following posts ....etc this type of data you don't need accounts you just need good proxies and the scraper will collect them without login, and this is will be easier, faster and cheaper of course because you don't need accounts, something to mention Business_email and email_in_bio could be extracted this way without accounts, while email in contact button should be logged in with accounts.

B. Data could extract ONLY with the logged accounts: data like followers usernames of the specific Instagram user could be only extracted with logged Instagram accounts.

So actually it depends on what you want to scrape ... in all cases, it's the scraper developer responsibility to use proxies or accounts/proxies based on required data type without Software user interaction.


3.Regarding your question about my software due to BHW rules, we can't discuss it in this section and the only valid place is the beta test thread which is here, if you ask any question there I will be happy to answer it ... I don't want you or me get an infraction from BHW staff.

Nice share. That said I'm doing fine scraping with ipv6 datacentre. To a certain degree anyway... Just not sure how much worth it buying 500 residential IP's is against 500 datacentre ipv6 (quite a lot more money)

Well, While I'm not a native English speaker But I'm sure I never said datacentre proxies always will not work, I said :

"I found most of the data center proxies detected and not good for IG, while most of the rotating residential proxies including 4 g mobile proxies are good and not detected by Instagram."

most does not mean all, So if you have datacentre proxies whether they are ipv4 or ipv6 and not detected that's really cool go with them, but if detected and still use them with Instagram accounts you should expect high banning rate, particularly for HTTP POST request like clicking like/follow/...etc button.

 

[rarexme]

Thanks mate for the share!!

 

[4gNET]

Well... I would like to know how do you think they know that you're using proxy. If you know how proxy works (especially CONNECT for https) then you know that it's not really possible to detect if you're using proxy or not (of course I assume you handle webrtc leak, dns leak, browser fingerprinting, etc). Moreover, I don't think they care that much if you're using proxy or not (correct me if I'm wrong but it's not even forbidden in terms of use), but it could be just one of the metrics to rate your credibility. What they care more is whether you automate stuff and that's way easier for them to guess analyzing behavior, do you download all resources, in which order, how do you handle all the stuff they have in this big obfuscated JS, etc.
So to sum up, they don't have single binary (true/false) metric which triggers ban, what they really have and do are multiple metrics (credibility of your email provider, credibility of country of your phone number you used for PV, IP location, phone number location, have you passed all IP leak checks, if you have disabled webrtc, which is not normal, etc). Using all this data they just select the most suspicious accounts and track them to gather more behavior data which they can use for ML to later easier rate/detect other suspicious accounts.
That's why it's also good to use your own private bot, it's easier to stay under the radar.

 

[thetrustedzone]

Well... I would like to know how do you think they know that you're using proxy. If you know how proxy works (especially CONNECT for https) then you know that it's not really possible to detect if you're using proxy or not (of course I assume you handle webrtc leak, dns leak, browser fingerprinting, etc). Moreover, I don't think they care that much if you're using proxy or not (correct me if I'm wrong but it's not even forbidden in terms of use), but it could be just one of the metrics to rate your credibility. What they care more is whether you automate stuff and that's way easier for them to guess analyzing behavior, do you download all resources, in which order, how do you handle all the stuff they have in this big obfuscated JS, etc.
So to sum up, they don't have single binary (true/false) metric which triggers ban, what they really have and do are multiple metrics (credibility of your email provider, credibility of country of your phone number you used for PV, IP location, phone number location, have you passed all IP leak checks, if you have disabled webrtc, which is not normal, etc). Using all this data they just select the most suspicious accounts and track them to gather more behavior data which they can use for ML to later easier rate/detect other suspicious accounts.
That's why it's also good to use your own private bot, it's easier to stay under the radar.

I agree with most of what you said but I think when you said "not really possible to detect if you're using a proxy or not" you mean residential proxy, right? because as you know with good IP/ISP database you could detect datacentre proxies ...

now back to the topic ... yes there are tons of factors they check for ... but the most important factor for them is Number and type and location of HTTP requests per time ... because mobile proxy IP shared between multiple mobile provider clients at the same time blocking IP based on the number of requests isn't the right decision for them ... this is why they applied more related factors ... for instance, mobile proxy belongs to London city mobile provider and sending high requests rate at 2 AM London time to french language/location profiles ... the high rate evaluation based on ML data they developed over time per location/language ... it's cat and mouse game ... related interesting article here (pdf document link ) from OWASP.

and yes definitely big social media companies can't disallow using proxy/VPN/tor ... in some countries with fascist political regimes protestors social pages and ppl following them need to use proxy/VPN/tor ... if big social media companies disallow that will show themselves as partners for that regimes ... it's human rights issue in the end.

 

[4gNET]

I agree with most of what you said but I think when you said "not really possible to detect if you're using a proxy or not" you mean residential proxy, right? because as you know with good IP/ISP database you could detect datacentre proxies ...

Then they're just detecting if IP is from datacenter, not if you're using the proxy.
You can use RDP session without proxy from datacenter VM and it will look the same as through datacenter proxy.
Here is the biggest advantage of mobile proxies. They can safely ban datacenter IPs (or even whole subnets) and they're sure that it wont affect innocent users. With mobile proxies, it's impossible to do.
Anyway... all I wanted to say is that they analyze so many factors that it's getting harder to do automation these days, especially with public bots.

 

[thetrustedzone]

Then they're just detecting if IP is from datacenter, not if you're using the proxy.
You can use RDP session without proxy from datacenter VM and it will look the same as through datacenter proxy.
Here is the biggest advantage of mobile proxies. They can safely ban datacenter IPs (or even whole subnets) and they're sure that it wont affect innocent users. With mobile proxies, it's impossible to do.
Anyway... all I wanted to say is that they analyze so many factors that it's getting harder to do automation these days, especially with public bots.

They deal both same datacetre IP as a proxy or on VPS ... both same for social media companies.
and yes mobile proxies is excellent choice compared to datacetre proxies ... this is what I'm saying !

 

[cyberminator]

The IG checker ain't working bro =D
{"status":"error","message":"An error occurred while sending the request."}

 

[IG Professor]

useful information. thank u

 

[thetrustedzone]

The IG checker ain't working bro =D
{"status":"error","message":"An error occurred while sending the request."}

If you are talking about this online tool http://datamining.shop/Home/TestProxy I just now checked :

but make sure to add your proxy without extra space, if getting an error page this means the proxy does not work for IG.

 

[cyberminator]

So.. let me get this right... the result would only "show your proxy working fine" and "your proxy is not working" ?

 

[thetrustedzone]

So.. let me get this right... the result would only "show your proxy working fine" and "your proxy is not working" ?

1. If the proxy working you will get "your proxy working fine !" in green color as I shared above.

2. if proxy Not working you will get "Your Proxy NOT working" in red color.

3. the third result you faced {"status":"error","message":"An error occurred while sending the request."}, because error happens and most probably due to authorization access if you use IP: PORT proxy and didn't add the tool IP-194.49.78.178- in your proxy dashboard you will get this error if you use IP: Port:username: password proxy and added the wrong password -like extra space- you will get this error.

 

[cyberminator]

So means that.. if the proxy is good and undetected by IG it will show "your proxy working fine "
If proxy is bad and is detected (blkacklisted) it would show "Your Proxy NOT working"

 

[thetrustedzone]

So means that.. if the proxy is good and undetected by IG it will show "your proxy working fine "
If proxy is bad and is detected (blkacklisted) it would show "Your Proxy NOT working"

yes , if you want could pm me your proxies and i will check them myself ... btw after around 48 hours -hope- the desktop client will be updated to contain the proxy checker.