Best youtube client?

[t1wt]

Question

Which of the clients is used by SMM panels for mass liking? Which of them has the least checks?

I went through all the web clients, but there are no options. It does not matter how ideal your requests are - the activity will still be written off.

But for example, the Android Music or ANDROID_TESTSUITE clients
ANDROID_MUSIC
IOS_MUSIC
GOOGLE_ASSISTANT
GOOGLE_MEDIA_ACTIONS

Could they have weaker protection?

Other details

No response

Checklist

• I am running the latest version.
• I checked the documentation and found no answer.
• I have searched the existing issues and made sure this is not a duplicate.
• I have provided sufficient information.

[LuanRT]

I think it's related to how each client handles attestation. Some of them don't do it at all which is why it just goes through.

[t1wt]

I think it's related to how each client handles attestation. Some of them don't do it at all which is why it just goes through.

I did it through clients without attestation.

For example, TVHTML5 does not have it separately for the like. But still, after 90+- minutes, a wave of write-off occurs.

Moreover, I tested each of the clients written above - and the same thing still happens. Perhaps I should make the context more complete on these clients. But in the case of TVHTML5, this did not help.

This makes me think that there is some alternative mechanism.

[Nansess]

I think it's related to how each client handles attestation. Some of them don't do it at all which is why it just goes through.

I did it through clients without attestation.

For example, TVHTML5 does not have it separately for the like. But still, after 90+- minutes, a wave of write-off occurs.

Moreover, I tested each of the clients written above - and the same thing still happens. Perhaps I should make the context more complete on these clients. But in the case of TVHTML5, this did not help.

This makes me think that there is some alternative mechanism.

from my experience with SMM panels some of these use privatized methods but a lot of the times, its just them mass making google accounts storing the cookies, and then have automated browsers open up and do it, with residential proxies etc

[t1wt]

I think it's related to how each client handles attestation. Some of them don't do it at all which is why it just goes through.

I did it through clients without attestation.
For example, TVHTML5 does not have it separately for the like. But still, after 90+- minutes, a wave of write-off occurs.
Moreover, I tested each of the clients written above - and the same thing still happens. Perhaps I should make the context more complete on these clients. But in the case of TVHTML5, this did not help.
This makes me think that there is some alternative mechanism.

from my experience with SMM panels some of these use privatized methods but a lot of the times, its just them mass making google accounts storing the cookies, and then have automated browsers open up and do it, with residential proxies etc

Yes, but SMM panels put likes without views.

And considering the speed of cheating, volumes it seems unlikely to me that browsers are used.

Still, it seems to me that somehow everything is done through requests.

[VORAPIS]

Hi
Changing client used will not influence how Like action works, this is a inherited operation,
Views are counted by Innertube only internally, all clients still use legacy api wrapper for tracking video interactions, as long as the automated bot is not loading HTML5/FLASH video player, it will not count the view most likely ^^

[t1wt]

Hi Changing client used will not influence how Like action works, this is a inherited operation, Views are counted by Innertube only internally, all clients still use legacy api wrapper for tracking video interactions, as long as the automated bot is not loading HTML5/FLASH video player, it will not count the view most likely ^^

But I'm not interested in views. I'm interested in the mechanism that SMM panels use to botting likes.

They botting without views, and likes are not written off.

[t1wt]

Hi Changing client used will not influence how Like action works, this is a inherited operation, Views are counted by Innertube only internally, all clients still use legacy api wrapper for tracking video interactions, as long as the automated bot is not loading HTML5/FLASH video player, it will not count the view most likely ^^

and don't different clients handle like actions differently? for example, web clients have BotGuard, while some android tv clients don't.

[LuanRT]

and don't different clients handle like actions differently? for example, web clients have BotGuard, while some android tv clients don't.

Attestation will only matter if your account is deemed risky by YouTube. For example, you usually don't have to execute the attestation commands for accounts that are old and used normally. Now if you do the same on a newly created account, the comment/like will probably be removed, or will only be visible to you, unless attestation is provided.

The Android TV client does have attestation by the way, it uses DroidGuard.

[t1wt]

Attestation will only matter if your account is deemed risky by YouTube. For example, you usually don't have to execute the attestation commands for accounts that are old and used normally. Now if you do the same on a newly created account, the comment/like will probably be removed, or will only be visible to you, unless attestation is provided.

The Android TV client does have attestation by the way, it uses DroidGuard.

It is there, but it seems not to be used for the "like" action. For example, TVHTML5 has attestation, but not for like.

The point is that I use pretty good accounts for my script. I use TVHTML5 client with perfectly complete requests. True, I use not quite honestly to obtain Bearer tokens: I take tokens from the official api v3. But I think it is unlikely that this is the case.

Perhaps it's the client_date parameter - I take it from my browser. But it doesn't seem to be used for detection.

Still, it seems to me that the problem is in some method of bypassing the likes write-off. Either some old client or some other method.

Perhaps I should add more requests for a real session (like playback, log_event).

Well, the last option would be to buy old accounts (2006-2007).

But honestly, I don't believe that quality accounts are needed. After all, some services botting up to 1 million likes.

Overall, I've come to the conclusion that likes are the hardest metric to botter.

[Nansess]

Still, it seems to me that somehow everything is done through requests.

Just because some services say they bot up to 1 million doesn't mean they do, lol. Most likes will fall off due to the quick number of volume or will take hours to complete. You know how much money is in that 1 million likes or 1 million view count nonsense? YouTube wouldn't let that shit slide easily. Also, back to you saying browsers don't make sense because it doesn't add views, YouTube won't add view counts to videos if they barely played the video. The automated browser action is so fast, YouTube doesn't bother even counting it as a view. Also, headless browsers are extremely fast. You just seem to not understand the concepts of these systems and how people build them. I use headless browsers consistently for some things, and the speed is outrageously quick.

[t1wt]

I understand the concept of this.

I have been using likes boosting for 5 years. My volumes are up to a million daily.

And an order of around 100k likes comes almost instantly. No server can handle such speed. Headless browsers may be fast, but not that fast.

And I tried it too. I used real account cookies, fingerprints, noise in canvas and webgl, and still the likes were written off even in headless mode. It seems to me that there is no point in emulating a web client - there is the highest level of protection.

But I may be wrong. How fast are your browsers? Can an average server with 128GB RAM handle up to 10,000 threads at a time?

I still stick exclusively to the version about using weakly protected clients and full automation on requests.

Think about it, my supplier botting likes at a price of $ 0.08 per 1000. And these likes even after 30 days retain 95%. Obviously, such a low price means minimal cost of resources.

And to avoid view, it will be enough to block watchtime requests. (in web version)

But I still have an idea. You can load a YouTube page (for example, YouTube music). And make a request for a like from a real browser. This method will make the request come from a client that must be real for YouTube (the presence of other requests). And the YouTube music client itself does not require viewing for a like, and allows you to like even non-music videos.

Anyway, thank you very much for your answer. I will try this method again.

Just because some services say they bot up to 1 million doesn't mean they do, lol. Most likes will fall off due to the quick number of volume or will take hours to complete. You know how much money is in that 1 million likes or 1 million view count nonsense? YouTube wouldn't let that shit slide easily. Also, back to you saying browsers don't make sense because it doesn't add views, YouTube won't add view counts to videos if they barely played the video. The automated browser action is so fast, YouTube doesn't bother even counting it as a view. Also, headless browsers are extremely fast. You just seem to not understand the concepts of these systems and how people build them. I use headless browsers consistently for some things, and the speed is outrageously quick.