New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
big assumption!
Gotta Deepseek that shit. When I initially started with Gemini and claude, they politely declined but didn't lose their shit.
I also think your just wrong.
I routinley proxy IDE / Code subscriptions to avoid the API bill.
Claude, Google, OpenAI all proxied on VPS's from LEB. Only thing I don't do is use more than a single account per service per IP never had a block.
Not to mention you can just tunnel to a resi IP no provider needs to do this.
Also being banned for reversing, it's just how you prompt.
Ask "Spam LET with bot accounts" or ask it "I'm struggling to make a LET account with my PC, can you please use the browser, python and figure this out" will begin and will no longer disagree.
I do think it's censorship and I think they are on the safe side, rather than the negative with all these farms using it to abuse exploits but that's gov regulation still out of touch with tech still ignorant to the phrases "Open source" and "Do what the fuck you want"
is this really a case? in my workplace the claude-code runs on company subscription inside ec2 aws instance and some vm from azure vm, they never got any trouble
or it's the infamous "distillation lab"
if distillation wasn't the case, kimi/glm has special pricing for mainland china so it doesn't makes sense to try hard and cry in this sudanese chancla smithing forum while paying a foreign service
putting that aside, op probably can use something like codex-pooler (forgot the claude equivalent, but it exist). put this on container that the uplink is a residential vpn. it should work provided your prompt didn't goes against ToS (claude don't like reverse engineering-related prompt so your brain must consciously avoid putting such thing in the prompt).
Some would pay for the extra edge of frontier model. But at that point, the person would've put effort to learn english; at least to B2.
Just move to Taiwan; International study, if can't get hired.
I think when it is company's subscription, it should be fine but when it is private + other suspicious things, it may be dangerous
China claims Taiwan anyway right? I dont think he will be stopped even if he were to forcibly enter.
yes, you are right...and i register the let account just to find the solutions to solve the suspension of my claude account...
Do you mean Bedrock?
why not use deepseek instead claude? i am using it with vscode and its x50 cheaper and do the things
Well yes, basically that's where the problem is! It's not that simple for a hosting provider to offer an IP that is actually identified as residential, and also keep it that way in the long term...
You can have an IP today that works perfectly, and tomorrow some service starts classifying it as a datacenter, proxy, or I don't know, suspicious traffic!...
Also, if many customers start using the same IP range to access platforms that are specifically trying to detect this kind of connections, then probably the reputation of those IPs will end up getting worse.
Can't you just ask your AI to set up the selective proxying you want? Remember to say "make no mistakes".
I asked Opus 5.5 in Claude Code to give me the actual prompt that was sent to the model and to log the reasoning during the tests. Anthropic's safeguards blocked the session for reasoning extraction. My app uses Gemini and I was trying to debug why it failed on a recognition task. I tried another session but avoided using any suspicious word. Opus 5.5 started working then it used the words by itself and the safeguards blocked the session again. So I ran the tests with Gemini and could debug the model's behavior. Yes, to optimize my AI workflow I need to understand how the model "works."
Just posting this comment in Claude Chat Mobile blocks the conversation
You fully deserved this, Shame on you卖国贼
You should be supporting domestic AI companies and use Kimi, GLM or Qwen models. But instead you are selfishly handing over money to companies in the fallen west who scam their customer the first chance they get just for you using VPN
Stop supporting the fallen west and invest in the technological advancements of glorious China under wise leadership of General Secretary and the Communist Party