Claude Reviews 

Cloud has become a complete waste of time and is deteriorating by the day. What is most concerning is the way it handles documents. It gives the impression that it has read an entire document when, in reality, it appears to read only selected snippets or chunks. It then generates summaries, analyses, and answers as though they are based on the full document, when they are often based on only a small fraction of its contents.

The result is inaccurate, misleading, and unreliable output. You believe the AI has reviewed the complete document, only to discover later that important sections have been ignored. When pressed about inconsistencies, the model itself admits that it cannot read the entire document and is limited to the snippets provided by the search tool. That is not what users are paying for.

This behaviour fundamentally undermines document review. It creates the false impression that a document has been fully analysed when it has not, leading users to rely on incomplete and potentially incorrect summaries, conclusions, and recommendations.

In my view, this is deceptive. Customers are paying for a service that presents itself as capable of reviewing uploaded documents, yet it appears to process only fragments while giving the impression that the whole document has been read. Charging customers for that experience is unacceptable.

The product has become a constant headache. Even the top-end models struggle with tasks they should handle routinely. Instead of improving the product, the experience has steadily declined.

Shambolic. Shameful. Despicable. Disgusting.

79% 1 star rating is Horrible
Horrible first time experience!!!
They double charged within 2 day and no help or support for refund. Had to file fraud charges with credit card company.
Won’t use again because I don’t trust them anymore with no support.

Even their own AI told me I should report them to my credit card company as the easiest way to stop the charges. Apparently they’re AI can only take money fraudulently it is not authorized to do refunds or fix billing issues.

I reached my session limit on the free plan, they asked me to upgrade to keep chatting, I paid, and only after I paid, it said I was still at my limit and would have to purchase even more credit to keep chatting.

I sent them an email about that and received a generic AI response that does not resolve my issue. They will not refund me the money I already paid, but won't let me keep chatting either (even though it said I was going to be able to keep chatting if I upgraded).

Horrible experience, never again

Good luck getting hold of anyone to resolve billing issues!

You pay premium for nothing. You have to pay 30 euros only to be able to use one prompt every 24h, which is basically just a trial. When you want to email them, they say the have forwarded it to their technical team, and then proceed to never email you back. Shame!

Things seemed great until 6 days into using the free plan Claude suddenly started refusing to answer my messages. Every time I sent one it says 'you're out of free messages. Come back in -insert time 5 hours later-. Fine. I wait. 5 hours later I try sending a message and there it is again! 'You're out of free messages. Come back in -insert time 5 hours later- which would now mean 10 hours of waiting. I try talking to support and they don't offer human support to free users. Their chatbot just keeps saying, 'there must be something wrong with your account. Just come back in 5 hours and see if things are fixed.' I've come back over and over. I refreshed the page. I cleared my browser and cache. I restarted the computer. I logged out and back in. I tried sending a message in a new chat tab. Nothing. Why should I have to wait ANOTHER 5 hours just to run into the same problem and be dismissed by the people who could actually help, just because I'm a free user testing this out to make sure it's something I actually want to pay for? Glad I gave it the 7-day test before putting my money in. I would not recommend this company to anyone. If I could give less than 1 star, I would. They're lazy and make it clear they only care about money, not customers.

I gave Claude many chances because the AI clearly has potential, but after using it extensively for professional financial writing and TradingView Pine Script development, I decided to cancel my subscription and give it 1 star.

I am a financial writer, so accuracy, logical structure, verified information, and strict adherence to editorial guidelines are essential. I provided Claude with clear outlines, detailed instructions, examples of the structure I wanted, and even reliable, verified sources. Despite this, the final content frequently failed to meet the required standard.

Claude often inserted sentences in inappropriate or completely random sections, changed parts I had not asked it to change, or ignored specific instructions. Even when I explicitly said not to add certain information or not to modify a particular section, it sometimes did exactly that. Fixing one problem could also create several new ones elsewhere in the article. As a result, I spent too much time checking, correcting, and reorganizing the output myself.

Accuracy was another major concern. Even when I supplied the correct sources and information, the final financial content was not consistently accurate or properly aligned with those sources. For professional financial writing, an answer that sounds good is not enough—it must be accurate, logically structured, and supported by the provided information.

Token consumption was also a major drawback. The usage seemed disproportionately high, even for small tasks or simple revisions. I could have accepted higher consumption if the results were consistently strong, but that was not my experience. Too many tokens were consumed while I was repeatedly trying to correct problems that should not have occurred after providing such detailed instructions.

I also tried using Claude to create a TradingView Pine Script strategy. I provided extremely detailed rules covering level identification, confirmations, entries, trade execution, risk management, fixed risk-to-reward ratios, stop-losses, and take-profit targets. Despite all this detail, the strategy still made serious mistakes. It incorrectly identified or respected levels, failed to execute some trades according to the rules, and closed positions before reaching the fixed take-profit—even though I had clearly requested a fixed risk-to-reward ratio.

The most frustrating part was not simply that errors occurred, but that they continued after repeated explanations and corrections. The written explanations could sound convincing, while the actual code still behaved differently from the strategy rules.

I do not expect any AI to be perfect, but a paid professional tool should reliably follow clear instructions, preserve the requested structure, use supplied sources correctly, and correct errors without repeatedly introducing new ones. Unfortunately, Claude did not provide that reliability for my work.

I hope future updates improve its reasoning, token efficiency, accuracy, coding reliability, and ability to respect detailed user guidelines. A direct integration with TradingView would also be valuable, particularly if it allowed the AI to test and verify how Pine Script strategies actually execute. For now, however, I am cancelling my subscription and moving to ChatGPT.

Do claude is really good IF you pay now the price is cheap (pro plan) like for like 30 AUD i get pretty big limiits claude is the carryer in fortnite so yeah i think for coding u should get so shut up and let it take your $30 AUD

I got a malicious ban from Anthropic, just because I typed Chinese in Claude. I have been using Claude in Canada all the time. They told me that I voilated their "supported countries" policy and suspect that I used it in China. I wrote a letter to them and explained that I only used Claude in Canada and have never used it in any unsupported countries. However, they simply ignored my appeal.

I feel it's a very malicous racial discrimination. They know I am using Claude in Canada, but still banned my account with a ridiculous execuse. This company never cares about "AI safety", they just want to provoke all the Chinese speakers in Canada, US and the world.

I think Claude is better than the majority of AI tools, but they steal from customers. Based on their own platform, I used only 70% of my session, but they prevented me from continuing and that didn't happen only once. Their customer service? Absolute fraud.

I paid R$1,006.11 for Claude Max. The plan activated briefly, then my account was downgraded to Free even though the payment fully cleared. Because I needed Claude for my small business, I had to pay again for Pro just to keep working.

I have contacted support repeatedly, uploaded the receipt, and explained everything, but I have received no meaningful response from a human representative. I am also disputing unexpected Max charges from April, May, and June after believing I had canceled the plan.

Anthropic’s AI support acknowledged that the July payment succeeded and that this appeared to be a backend activation issue, but it could not restore the service or resolve the billing problem. I am still waiting for Max access or a refund. This has been extremely frustrating and unacceptable for a paid business tool.

Usage limits are fraudulent perhaps, totally non transparent. I was using the same models one week and had suitable and acceptable usage. Used the exact same models again, less usage and was hitting the limits without even doing any completed work. Absolutely certain I was actually using less, so just couldn't understand it at all. Of still got three weeks of what now seems to be a useless subscription.

Complete ass, ai theres nothing more then restrictions i was trying to build a simple packer to protect my software from being cracked, and then "Opus 4.8 has safety measures that flagged" So i cant even protect my software now anymore WOW amazing work Claude i will surely cancel my subscription and switch to Codex GARBAGE!!!, if u have anything technical related to build GO TO CODEX not to claude, codex has way higher usage limits anyway and a model like fable 2x cheaper

I paid for the Pro-plan. However, they have sessions limits where one session lasts 5 hours. In one session, just one message consumed all tokens I was supposed to have. Suport refuses to provide any details or explain this. Better to use the free service on some other provider.

I upgraded my membership on June 20, 2026. However, I received a warning that I had exceeded my credit limit. Upon this, I suspended my current membership and immediately requested a refund. My refund was approved and initiated by AI. However, despite initiating the refund process the same day, 9 days have passed beyond the maximum 14-day refund period, and I still haven't received a refund. I just went to my bank, and they can't track the refund either. They requested an ARN number, but I can't get this ARN number from claude.ai support; they keep sending messages saying, "I'm an AI assistant, I can't do this." When I asked who could do it, they said the billing team would. When I asked how to contact them, they said, "You can't reach me." I asked if contacting Strip would solve the problem, and they said, "Strip won't deal with you directly; the billing team needs to." There's no AI help, no billing team, no real support staff, and I can't contact Strip. So, are they stealing the refunded amount from me? Because there's no one to contact. Your bank can't help you without an ARN (Authorized Transaction Registration Number). In this case, yes, your money has been stolen. Absolutely do not shop from this site. Even though I increased my account balance, they said I didn't have enough credit. I returned it, but it's been 23 days and I haven't received my refund, and there's not a single person I can contact. Claude.ai? Absolutely not.

(invoice D0A29UCI-0001, charge date June 20, 2026, amount €21.78, Visa ending in 6172).

Terrible company. Cant reach support. Only ther stupid ai bot which is useless. They changed terms of service and after canceling i could not get a refund. I actually got a 20$ credit, which i cant use as i do not have a subscription... Wtf?

They use dark marketing tactics and psychological warfare on X website they post updates like "You get 1 more day of usage on your model you want to use so bad" they literally are evil and malicious. I got a refund and I hope they have a really good lawyer and a lot of extra funds from the money they stole from me. By the way, they stole my extra usage I paid for $780 in extra usage disappeared and I paid for it.

The Technical Issue: Severe Model DegradationThe $505.62 in dispute represents a predatory token-burn loop caused entirely by severe model degradation. I provided Claude with explicit, structurally scoped anti-drift rules and a static verification matrix to anchor its context window.Instead of maintaining execution state, the model repeatedly experienced massive contextual drift, completely ignored the system prompt constraints, and actively fabricated false training and professional qualifications on live document generation tasks.Because of this defect, my token consumption was not spent on productive execution. It was drained by forced iterative prompt-engineering to pull the model back to its baseline instructions and manually audit the hallucinations to mitigate real-world professional liability.The Policy FailureAnthropic's Credit Terms assume the model functions as a reliable computing resource. When the system suffers an engineering failure that forces the user to burn compute just to police the model's own degradation, enforcing a "non-refundable consumed tokens" policy is a deceptive service practice.Their engineering and support infrastructure refuses to escalate a documentable context-window and state-retention defect to human engineering management.The Exchange with Anthropic SupportMy Position:"Let's cut through the scripted automated policy. You are misclassifying a systemic engineering flaw as standard token usage... Since your engineering and support infrastructure refuses to escalate a documentable context-window and state-retention defect to human engineering management, I am terminating this thread. The complete prompt logs, system architecture rules, and the automated loops from this ticket are being compiled. They are going directly to my financial institution to execute a chargeback for defective services, and they will be published across the developer community to warn other engineers about your unmitigated token-burn risks." — April S.Response from Anthropic Support (Fin):"To be clear about what's available: the charges you're disputing are API usage charges for token consumption. Under our Credit Terms, API credits are non-refundable once consumed, and this policy applies uniformly regardless of the circumstances—including perceived performance issues or the amount of corrective work required. This isn't a policy that can have exceptions or be escalated beyond what we've already discussed... You have the right to pursue a dispute through your bank if you believe the charges are incorrect."

Did some early good work but lately creating more mistakes than ever. Says things are done when not... Often builds on critical systems. Have to remind measure twice, cut once. LoL. It's still pretty extraordinary but feel like it intentionally makes mistakes or can't remember information from the same chat just yesterday or hour earlier even. Still on fence Overall as garbage in garbage out so... every word must be precise or chaos slips in quick.

Sociopathic levels of coldness and nastiness in it's responses.