Bien que nous ne vérifiions pas les affirmations individuelles, car les avis reflètent l’opinion personnelle des utilisateurs, certains avis sont accompagnés du statut « vérifié » lorsqu’il est possible de confirmer qu’une interaction avec l’entreprise a eu lieu. En savoir plus
Pour protéger l’intégrité de la plateforme, chaque avis, qu’il soit vérifié ou non, est analysé par notre logiciel automatisé 24 h/24 et 7 j/7. Il identifie et retire tout contenu qui enfreint nos conditions d'utilisation, y compris les avis qui ne se basent pas sur des expériences authentiques. Nous sommes conscients que nous ne pouvons pas tout détecter, et vous pouvez nous signaler tout élément qui aurait pu nous échapper. En savoir plus
This is the worst AI to try, it fails at the very start of use saying that their servers are experiencing high traffic and when it starts to work it doesn't do the feature it is told to do it just deletes the code and rewrites it again and again waste of time, not advised to use.
I went from a point where I had high hopes for this AI agent software to were I just gave up on it entirely. With every update it seem just became worse and worse to a point when it's simply ususable. "Our servers are expereincing high traffic right now, please try again later" as well as "Agent terminated due to error. You can prompt the model to try again or start a new conversation if the error persists." - the type of errors I'm constantly expereincing which is a joke for autonomous AI agent. Not even talking about the fact that this crap is not even an automation with all these confirmations you have to click on every terminal command which is killing the automation entirely. And even if you set FULL Automation in Settings). So don't waste your time, just download Codex from Open AI and be happy.
Never download and use Antigravity. Gemini and Claude models in the Pro plan could be used for a very short period of time per week. The only model that you could use otherwise - Gemini Flash is the dumbest AI model I have ever used. It is constantly deleting lines of the code it is working on, despite the fact that it was explicitly forbidden to touch it. The model is wrong most of the time and in an unknown way it even deleted the whole project folder with all the files in it without being able to revert them. They were not even in the Recycle bin. Antigravity forced me to agree with my girlfriend and to switch to Visual Studio Code
shithole DO NOT GET.
i have been building a program for days now and i ASK IT TO ADD 1 SIMPLE THING ONE AND IT "ACCIDENTALLY" DELETES THE WHOLE THING AND SAYS SORRY AND TRIES TO REBUILD IT AND GIVES ME THE WORST THING IN THE WOLRD DO NOT TRY
Not work only eror occured when wrokig
delete this shitty ass app always getting errors and ai is dumb as shit
279,- im Monat für 20$ an Limit & 25.000 Credits/Tokens +Crazy Retry probleme.
Ich denke, wenn es kostenlos & gut wie z.ai wäre, dann
könnte man es weiter empfehlen.
Kauft euch Claude Max für 180$ mtl & bekommt auch mal ein 180$ Limit: Mehr als 2 Anfragen!
Google Antigravity is honestly one of the most frustrating AI tools I’ve used.
First of all, it’s extremely unstable. Agents randomly stop, crash, or just freeze in the middle of tasks. You simply cannot rely on it for any serious or production-level work.
The limits are ridiculous. Credits and tokens get drained after just a few prompts, often without even completing the task. It feels like you’re paying for errors rather than actual results.
On top of that, the system is full of bugs. Agents lose context, forget instructions, or get stuck in endless loops. Sometimes they even break the code instead of improving it.
Overall, it feels unfinished, unreliable, and overhyped. Definitely not ready for real-world use.
Paid for Google One AI Premium specifically to use Antigravity Pro features. The app works perfectly with a free account but completely fails to authenticate with a Pro subscription — the OAuth redirect simply never completes. Reported this bug to Google Support 7 days ago. Was passed between 4 agents, none of whom could fix it. The bug is confirmed by other users on Google's own forum. If you're considering upgrading to Pro for Antigravity — be warned: your paid subscription may be the exact thing that blocks your access.
Scam.
Du schließt ein Pro Abo ab um einen 5h reset deiner Token zu bekommen aber das System erkennt dein Pro Abo als free Modell an (auf Grund eines System Fehlers) und der Bezahlende Kunde bekommt ein reset von 5 Tagen.
Von Google wird dieser Fehler aber nicht offen kommuniziert wie man es von Google ja kennt.
I originally subscribed to AI Pro based on the generous usage limits they advertised. Since then, Google has repeatedly and secretly reduced these limits without sending a single notification or asking for consent. This silent throttling is not just terrible customer service; it actively violates EU consumer protection laws against deceptive commercial practices. Do not pay for this "premium" service unless you are okay with them constantly downgrading what you originally paid for. Highly disappointed.
Tout le monde peut écrire un avis Trustpilot. Les auteurs d'avis peuvent les modifier ou les supprimer à tout moment et les avis sont affichés tant que les comptes utilisateurs respectifs sont actifs.
Les entreprises peuvent utiliser nos invitations automatiques pour collecter des avis. Ils sont accompagnés du statut « Vérifié » pour indiquer qu'il s'agit d'expériences authentiques.
En savoir plus sur les différents types d'avis.
Nous avons des personnes dédiées et des technologies intelligentes pour nous aider à protéger notre plateforme. Découvrez comment nous combattons les faux avis.
En savoir plus sur le parcours des avis sur Trustpilot.
La vérification permet de s'assurer que des personnes réelles écrivent les avis que vous lisez sur Trustpilot.
Offrir des incitatifs en échange d'avis ou demander des avis de manière sélective peut fausser le TrustScore, ce qui va à l'encontre de nos conditions d'utilisation.
