Anthropic Reviews 216

I was using claude 4.6 Opus and it was great, after launching 4.7 they have dropped 4.6 intelligence and it can't even handle a simple issue, it seems they dropped 4.6 intelligence to 4.5 and renamed 4.6 as 4.7, I don't wan't to pay for 4.7 while it is the same as 4.6!!! i just want the old 4.6 one back!!!

What the actual f-uck of service? Simply just lobotomizes their models. I set claude cli on max and it is dumb as smallers models without thinking. What is the point of paying money for shitty service

I redeemed a gifted Max plan on April 9, 2026 (receipt #2672-1740-8629, valid until May 9). It worked fine on April 16. On April 17, my account suddenly showed "Free Plan" with no access to Max features.

I contacted support immediately. The Fin AI agent confirmed it's a backend issue but admitted it cannot create tickets or restore access. After multiple attempts, I'm stuck in an endless loop of generic suggestions that don't apply to gifted subscriptions.

There is no way to reach a human agent. Days of my paid subscription are being wasted. This appears to be a known billing synchronization bug (similar to GitHub issue #31012).

Anthropic, please manually restore my Max plan. I have provided all documentation: receipt, invoice, screenshots, and account email.

I will update this review once a human agent resolves the issue.

I subscribed to Claude Pro Annual on April 12, 2026. Two days later, on April 14, Anthropic charged my card twice — $100 each — for a product called "Gift Max 5X" that I never purchased, never authorized, and never used. A third charge of $200 was also attempted on April 17 (failed because I had already blocked my card).

These transactions appear in my own billing dashboard. I did not initiate them.

I spent an entire night going back and forth with Anthropic's "Fin AI Agent" support bot. The bot:
- Offered to refund my annual subscription (which I didn't want refunded)
- Directed me to check "audit logs" that don't exist for Pro accounts
- Repeatedly redirected me to my bank instead of taking responsibility
- Never escalated to a human agent despite multiple explicit requests

This is not an isolated incident. Other users on Trustpilot and Reddit have reported identical unauthorized "Gift Max" charges around April 9–12, 2026. This is a known system bug.

I am now proceeding with a formal bank chargeback. I have documented the entire support conversation as evidence.

Anthropics product is excellent. Their billing system and customer support are a serious failure.

Les meilleurs , ils ont mis une ligne droite à tout le monde

I got a weird email about "suspicious signals" and hence that the account was being suspended. I used this account for only a few days and there was absolutely nothing suspicious or improper in my use. There is a general form to complain/inquire about this. Emails lead to an automated ai helper. This agent responds asking how it may help. Seconds later (before one can answer its questions as how it may help), this ai agent sent another email providing the same general trivia in the initial received email about the suspension and then moments later after I attempted to ask what is going on concludes with a third automatic email that the case is closed.

Worst after sales service it’s a shame for a company that try to digitalise everything, I’m really disappointed

Abonnement payant limité comme une version gratuite + manque total de transparence

Je suis cliente avec un abonnement annuel de 216 € et je constate aujourd’hui un dysfonctionnement majeur : le service est inutilisable dans un cadre professionnel.

À plusieurs reprises dans la même journée, mon accès a été bloqué après environ 45 minutes d’utilisation, avec impossibilité de reprendre avant plusieurs heures (jusqu’à 23h).

Ce type de limitation correspond à une version gratuite, pas à un abonnement payant.

En pratique :

-interruptions brutales et répétées
impossibilité de travailler de manière continue
perte de productivité directe

Le point le plus problématique reste le manque total de transparence :

-aucune information claire sur les limites réelles de l’abonnement
-aucune anticipation possible de ces blocages
-aucune explication fournie

Et surtout : aucun accès à un support humain

-uniquement un chatbot automatique
-réponses génériques
-impossibilité de résoudre un problème concret

Dans ces conditions, il est légitime de s’interroger sur le positionnement tarifaire, d’autant plus que des services concurrents comme OpenAI proposent une utilisation beaucoup plus stable et continue, sans blocage après moins d’une heure d’usage.

Cela donne le sentiment que les limitations sont volontairement restrictives, sans communication claire, ce qui pose un vrai problème de confiance.

J’attends :
-des explications précises sur les limitations appliquées
-une mise en conformité du service avec un usage professionnel
-ou à défaut, un geste commercial adapté

En l’état, le rapport qualité/prix n’est pas acceptable pour un service payant.

I subscribed to Claude Pro expecting a functional AI assistant I could use throughout my workday. What I got instead was a series of undisclosed restrictions that I only discovered by hitting them.
First, the token limit. It runs out within minutes of normal use — not heavy use, just regular back-and-forth work. Then you're locked out for five hours. This is nowhere on the pricing page.
I adjusted. I restructured my entire schedule around these 5-hour cooldown windows to maximize the time I actually had. I accepted this as a compromise even though I was paying for a service that was available to me for a fraction of my working hours.
Then I discovered there's also a weekly cap. On top of the hourly limit. On top of the cooldowns. Completely unadvertised. No advance notice. Just suddenly locked out for days after following their own rules.
This isn't a product limitation — it's a transparency problem. If these limits were clearly stated before purchase, subscribers could make an informed decision. Instead, Anthropic takes your money and lets you discover the restrictions through frustration.
I've contacted support multiple times. I'd encourage anyone considering a subscription to understand that you are paying for extremely limited access that is not clearly communicated.

Utterly disgraceful service, as has been outlined by plenty of others.
Customer service is non-existent.
Bought their Claude Pro Plan to assist with work. I am constantly reaching 'limits' that basically stop me using the service for several hours. In one case, it was less than an hour of operation before being cut off.
The final straw came when I was looking to generate a list of 50 email addresses. It promised it could do it. And then didn't. Repeatedly. And then gave up and said it couldn't do it after all.
Chats fill quickly, and slow the page down to the point where it's minutes between messages. It then stops completely, and you have to start a new chat. While it says that information committed to memory can be recalled, every single time I've had to start a new chat, it simply cannot follow what was in the previous chat, so you're basically starting from scratch.
I could go on for hours about how useless the service is, and you can't even get any help from their non-existent support. I tried reaching out to their customer service and after trying to get past the irritating bot, it said a live person would join the chat. 37 hours later, no one had.
Their policy is refunds for 14 days, but it's impossible to know what you're going to need in those 14 days and, as they charge you for a full year, they won't even pro-rate the refund. I'm fine paying for time used, but if I cancel, I shouldn't have to pay for unused months of service.
Disgusted. Avoid.

Tous les jours un bug ya qu’à voir les serveur status c’est nimp

They must give a free subscription to users whose subscription is not working.

Just ai nonsense response to a billing problem from Fin, no help to get. I lost credits when I renewed my account, so this should be a human involved. I don't think they have any in support.

CC was good enough in march but now I have noticed a huge degradating of the Chat bots logic, makes mistakes far more often and wastes you tokens because of it.

opaque functionality, noticeably degraded product, but still charging full price

On April 12, 2026, Anthropic charged my debit card twice for $238.00 each (total $476.00 USD) for a product called "Gift Max 20X - 1 month" that I never requested, never authorized, and never used. I was on a Claude Pro Monthly plan at $23.80/month and never changed it.
The same night, they attempted a third charge of $238.00 and a fourth of $119.00 — both failed because my bank blocked them. That's 4 unauthorized charge attempts in one night.
I have two separate receipts with different numbers (C1BF5803-0021 and C1BF5803-0022) proving two real transactions. My billing history in claude.ai confirms both charges.
This is not an isolated case. Another user from Taiwan reported the exact same unauthorized "Claude Max 20x gift" charge on April 9, 2026 — just 3 days before mine. This is a known system bug affecting multiple users.
I opened a support case (ID: 215473880288650). The first response was an AI bot that offered to refund only $23.80. After providing full evidence, my case was escalated to a human agent, but no resolution yet.
Anthropic must investigate where these charges originated, who initiated them, and from which IP address — because it was not me.

Messages bettwen me / ai are often deleted for no reason, raised a complaint with Anthropic about this,no response. Poor customer service
Used to be a paying customer, not any more

Well I joined in March and loved it. Editing a bok chat gpt kept taking forever to help me with.. Well April 1st that all changed. I loginto claude on my laptop and 20 min in, right after reuploading my book cuz claude doesn't save like gpt, I get a message saying I used up all my time,, TWENTY MINUTES IN! As soon as I uploaded my book, time used up. at that time they claimed there was a system outage . Well just tried tonight, same thing.. the first time I got an agent that didnt do a damn thing, tonight bot refused a human saying it wasn't an option till I pointed out they are in complete violation of consumer protection laws in WA state, they finally said they would send me to a human, well almost 2 hrs later, no human response. And Now Im way past my books deadline..

Claude au was great, better than chatgpt, but then my account got suspended because "A child used my account", I am not a child, and I am NOT going to freely give my information away to a company which is gonna sell it to every other company in existence. Never again am I gonna consider using any of anthropic's products, EVER AGAIN!

After i changed plan they wiped out my api tokens. Complained a week ago, still zero response. Either dishonest or terrible service, you choose.