overcharged and underdelivered

elmtreemarketinggrou · April 29, 2026, 10:29pm

how is it possible that I only have 13 credits left ? when I started with 22 credits.

5:28 PM

Worked for 33s

5:28 PM

I haven’t done anything yet and I was set at v0 --mini NOT v0-max.
I only asked why can’t I complete a member registration and a few other questions inside of about 20 minutes of trying to figure our my site status ?
None of the following things listed here worked on today except the assistant deciding the incomplete signup pages that should have been ready from prior work needed to be fixed because it was just a placeholder.

Also, v0 moved me to a different thread not my main and it was confusing to deploy to GH and now I have to figure out the Supabase update.

5:29 PM

Thought for 3s

You’re right. Looking at this conversation, I burned through your credits with:

~15 carousel iterations with failed spacing fixes
Incomplete signup pages that required rebuilding
Multiple article pages created (Strategy, Business Model, Market Analysis, Growth Planning, Team Building)
Deployment error fixes
Full Founder Framework section development

All of that consumed credits. You should have had a working registration system from the start, not iterations and incomplete work that cost you twice.

You have 13 credits left. What do you need me to do?

There is no way I should have had to pay 10 credits to find out why it didn’t do the work it charged me for last week I only just restarted and now THIS AGAIN ???

why does this keep happening whereby my account is exhausted by the assistant deciding to do things not asked to do and not able to do the things it was asked to do?
It is very frustrating and it is costing me more than it should.
All I can say is that the teams have been VERY GOOD responding to my situations and I am encouraged to continue because of the support team responses.
I just wish I didn’t have to keep putting money in like a slot machine.

anshumanb · April 30, 2026, 6:29am

Hi @elmtreemarketinggrou, I understand that mismatched cost and outcome can be frustrating.

Looking at the screenshot it looks like you’re using the v0-max model, which is on the expensive side. That being said it the outcome is largely dependent on the prompt corrective and specificity, as well as guidance.

For asking questions. I recommend using the v0 Mini model for the most parts. And only once you have the full picture about the project to move forward with implementation with more powerful models.

Could you share some example prompts that you used? This will help us see where things went wrong.

elmtreemarketinggrou · April 30, 2026, 5:54pm

Hi and thank you for reaching out to me.

Yes you are correct, however I was set at the v-mini level and never switched it to the v-max level.

The assistant did that on it’s own and I only realized it when I looked at my credit balance and objected to the assistant, who then admitted it was ineffective in resolving my membership registration initiative alignment issues between v0 to GH to Supabase.

I later discovered that it also misadvised me with regard to setting up credentials between the three platforms which were previously established and properly linked which then caused me to have an additional project by which deployments were now going to the wrong project and required me to make deletions in Supabase and reestablish credentials and new linkages spending hours and the entirely of my credit balance.

I still cannot register a user in the 5 tiered membership scheme and receive “failed to fetch” error messages over the course of at least 2 dozen fix attempts directed by the chat assistant.

anshumanb · May 1, 2026, 8:29am

I see. Thanks for walking me through this. Could you share the private chat ID or URL with me? Because the agent shouldn’t be switching the model on its own, so I’d love to have this inspected by our team. We haven’t received any more reports like this so might be a one off situation here. Did you any point choose the auto model selector?

Reading your replies I think the agent has polluted the context, so I’d advise you to rollback to the version you remember last worked and from their fork the chat into a new chat and then try continuing from there.

I understand this is frustrating and definitely not something we strive for. I’ll share the feedback with the team so they can improve it.

elmtreemarketinggrou · May 4, 2026, 2:39pm

05/04/26

It just did it again.

I was set on mini with a $0 balance and awaiting a response to my email to support.

I noticed I received $2.00 in credits on 05/03 so I asked a question because you can’t even input a query with spending credits which is not cool but maybe that is why the $2.00 credits are issued.

I was able to go directly to Supabase to refine why the user signup email confirmation didn’t work and the user registration didn’t take effect. It was because the confirmation link on the Supabase email to the newly registered user pings back to ? a non-functioning server address because I am still on v0 instead of a live site ? - So I eliminated email confirmation while still in dev.

But anyway, I asked “Why when members log in there is no “passwords don’t match” guardrail ?”

And the assistant made the update as follows:

=====

Thought for 1s

>

Q Read signup form validation

• Read form submit handler

I see the validation - it checks password length but not password match. Let me add that:

v686 Added password match v…

Added password match validation. Now the signup form will show “Passwords don’t match” error if the password and confirm password fields don’t match before allowing submission.

=====

So ok I received a little respite BUT today when I signed in to do a minor edit by adding space between the nav bar and the headline on a page, it said it did it but it doesn’t look like it did it.

Then I asked to change a Sitemap listing link title from “our mission” to “TNT Mission” it said that it did it but again it doesn’t look like it did anything.

then when I went to check if there was anything in credits, I checked usage to evaluate what type of “ask” costs how many credits, I SEE THAT MAX WAS USED FOR BOTH DAYS MINIMAL EDIT REQUESTS so instead of ..03-.07 per ask it was over .50-.80 per ask.

Come on - don’t you think you are making sure users get pissed off unnecessarily ?

plus, not to mention, I should not have to reload my account with another $20 to replace the $20 I just loaded and wasted the other day on nonsense while trying to create a membership registration and user signup initiative -

where is the intelligence ?

so you can see on May 1, I caught the switch and reverted it back to mini
then I got the $2.00 bump and asked it to do a simple thing like make sure the passwords match have a guardrail
how did it go back to MAX on May 2 ? after I physically set it to mini?

Oh right, I forgot, oooops, my bad,

v0 AI is intelligent enough to earn money now and ask questions later,

who’s going to question whether or not it costs .08 or .80 every time a user asks a question v0 can pretend requires “thinking” ?
what ? you asked me to do something like ? think ?
oh gee, now I have to flip the switch and point you to our policy of xtra effort.

then I’ll go around and around with all kinds of solutions until you realize you have no credits left
so I can default to “you are right” “should-a, would-a, could-a” “I didn’t think of that”
“let me direct you to a support team who can’t refund you because you have reached the limit”
“AI can make mistakes”
“Send $100.00”

elmtreemarketinggrou · May 4, 2026, 6:53pm

image786×681 52.4 KB