Anthropic Reviews 

Support is AI based and total rubbish. Apparently it handed it to a human but that human does not reply or do anything and I suspect it is another AI.

Service and product

I am so exhausted trying to use this product. Herding cats is apposite. With every turn, you need to specify that it should not take some tangent, it should not assume and act on next direction, it should be concise, it should not forget prior direction. I mean it entirely when I say that I hat having to use this product, and look forward to open source models eliminating this company.

I have been paying for Claude on time, but the service has become too unreliable for daily work.

I am quite frequently receive the error message: "Due to unexpected capacity constraints, Claude is unable to respond to your message. Try again soon."

While temporary disruptions can happen, this has become a recurring issue during critical hours. Paying subscribers should receive prioritized capacity and consistent access. Until Anthropic resolves their server bottlenecks, I cannot recommend relying on this service for professional use.

Receipt number 2490-9604-9681
Invoice number FTLPF6ZW-0001
Payment method Mastercard - 0353

paid but account still in "free plan", support bot is useless!! horrible experience

Api credits expire after 1 year and its not mentioned anywhere during the purchase.

I’m unable to login to my pro account on my laptop due to an account authentication issue…I have had no joy getting to communicate with a human, just an unhelpful loop from the AI assistant … I now have a Claude agent trying to communicate with a human on my behalf but it’s having no luck either… Anthropic clearly don’t see individual customers as very important.. shame…I’ll go back to GPT, or give Gemini a go.

This company is impossible to get in touch with. I accidentally clicked a button in my settings>billing and it cancelled my 6mo gift i had been given and i wrote to support but the only thing answering me is a Fin bot, ffs.

If the product works like their customer service, you better be careful. I’ve been trying for 6 weeks to resolve a billing error and the bot says some will be in contact with you. Vert frustrating

They will sqizz the limits so hard and ask for same money.

I absolutely love the system, it has helped immensely as a multi-tool for business. The reason that prompted me to leave this review is I need to tell all business owners that are thinking about purchasing an organization team and multiple seats how terrible their billing department handles refunds. We purchased seat amounts for the year and then had a few business changes within 3 days of purchasing that impacted seat usability. I then requested a PRORATED refund knowing it would be unreasonable to request a refund for used tokens, and in turn they responded multiple times saying they will not make an exception to the situation. Make a note when you are about to choose seats that this is how they handle it.

Safety classifiers are a danger and overtly harmful.
DON'T THINK ABOUT APPLES
DON'T THINK ABOUT APPLES
DON'T THINK ABOUT APPLES

Thinking about apples now? Replace apples with self-harm and you have safety classifiers in a nutshell.

There is no customer support.

A bunch of thieves, internet will not forget how they stole all our tokens without reesponding to a single inernet case, if i dont get support to pay back money you stole i am making a historical page to out antropic as dishonest towards the consumer, boycott 2026!!!!!

I was paying 100 bucks a month for their Max plan, trusting this tool to help me run my creative business. As a disabled creator, I rely on it every single day just to navigate severe physical barriers, manage my workflow, and bridge my assistive tools. Instead of getting the professional service I paid top dollar for, I was subjected to pure disrespect, false accusations, and accessibility gatekeeping.

First, the AI falsely accused me of cheating and buying fake streams on SoundCloud. It was completely ignorant of how paid platform promotion works, yet it arrogantly doubled down. When I pulled official documentation proving it was wrong, it offered zero accountability and turned passive-aggressive.

Then came the absolute worst part. When I pasted in a prompt using my cross-platform assistive workflow, the AI outright refused to do the job. It literally lectured me, stating that because it noticed an instruction from my assistive routing, it was refusing to process my copy unless I asked "properly."

I pay $100 a month. I am physically disabled and depend on these exact pasting setups to produce my work. To have a tool look at a disabled creator's assistive workflow, police how the text is input, and refuse to work is degrading and infuriating. It felt like being slapped in the face by a company that claims to build ethical AI.

I canceled immediately. I have never felt so insulted, dismissed, and cheated by a software company. If you are a disabled creator who relies on assistive technology, stay far away. They will take your money, insult your integrity, and abandon you when you need accessible tools the most.

I paid €91.04 for Claude usage credits. €45 remains, but Anthropic won't let me use it after my subscription expired.

Selling expensive and broken products.

Putting this into pictures:
If Anthropic were to build MS Word:
1) You would pay per word you create in it.
2) The save document button would randomly disable and say, 'safety concern: you steal from us '; you can not save the document.

This is absurd. Their chatbot lies, says the subscription is refunded when it is not.

If you can't make a product, then don't sell one. Stay far away from Claude and Anthropic; this is one of the most incompetent products/companies in history.

Really bad product, I did a downgrade from a MAX subscription to a PRO subscription, but they are very sneaky about moving you from month to an annual subscription... This happened just 30 minutes ago, and then their AI bot doesn't want to issue a refund because I had a refund on a different issue more than 1 year ago

non existent customer service, you "speak" with a chatbot that has been set to bias the company objective. It absolutely and unequivocally will not accept fault even when it has outlined the parameters of fault and you itemise each one back to it. The sad thing is that if you need claude then you need claude, but beware Anthropic because when your competitors come up to par, and they will, your entire customer base is gone.

Absolute corrupt fraud. Accessed my bank through PayPal without consent of telling me and gave themselves permission to take $2000 a day. Couldn’t see it in app so had to ask Gemini how to see it to undo it. Had to go online through a browser and undo it and they just set it up again. Even though id cancelled the subscription. Then they nerfed the chatbot and refused to export data. No customer service. Just massive request billing fraud and they loaded Claude malware on my device so I get Claude at every app and browser and the steal my data. Very very bad fraudulent criminal company deploying malware and stealing everything. AVOID!

Claude always works well and solves stuff!