Yes, totally understandable.
itās definitely on the roadmap, Xano is working on their side to enable their services to be built / triggered by third-party AIs, my understanding is that they should be ready towards the end of March. Once they are, weāll build the AI integration.
1 Like
I still see the On mounted hook
Checking right now
Hey @fbascunan and @pitt2k, so we just added a āin-betweenā hosting add-on called āGrowā with 200Gb of bandwidth. So for those of you in need of as much bandwidth as with the legacy starter plan, you can migrate to this plan: https://pricing.weweb.io/
Thanks again for the feedback, hope that helps!
1 Like
Hello @Dorian, I confirm that now tokens will not be counted when the AI returns an error.
Hi @AgentD,
Thank you for your kind words. We didnāt plan to remove this trigger, so I completely understand your frustration that itās missing on your end.
To ignore
I can still see it on my side, so Iāll need more details to understand whatās happening with your project. Could you reach out to the support team here 
, and tell them on which element you have this issue? Iāll make sure they prioritize your request so we can investigate.
I just checked with the support team, and it was fixed an hour ago. Can you confirm if everything is back to normal on your end?
1 Like
Excellent!!! Iām developing tools for internal use, and I think that is the perfect price point for my clients, considering the use. Thank for the very quick response!
1 Like
Great, thanks for the feedback!
@Raphael I am still seeing this issue. Any idea on ETC please? Thanks.
even if the AI was working, it would still run out far too quickly even on the paid membershipsā¦from my experience of getting to 1M token doing hardly anything.
Yes @flo thanks a lot guys. It is all good now, the triggers On Mounted and Before Unmount are both back! 
2 Likes
This 3.0 version got me excited. Back in October I put on hold the project I had been working on for a few months, partially due to capacity and partially as a result of lack of confidence in my execution of WeWeb + Xano.
This release gave me some energy to see if it would help fill the gaps.
Wellā¦my goodnessā¦I have rebuilt from scratch everything I had previously built (Yes, started with a new project) and FAR more in a fraction of the time. But not only that, the UX is far superior than the previous build I had pulled together.
An example is the rollover nav item effect. Could I have done this pressing all the buttons? Probably, with enough time, but all I really did was describe what I wanted, made some edits after the AI produced the results, and tweaked a couple details manually myself.
Effectively Iāll be ready to invite some of my contacts to try out what Iāve producedā¦within the next couple weeks.
Had I been using Supabase perhaps it would have been even faster. But Iāve stuck with Xano and have become proficient enough to get what I want in most cases.
All this to say Iām blown away with the outcome here within 5 days since giving it some time, likely less than 15 total hours. That is mind blowing, all without being a back-end or front-end engineer/developer.
Incredible work to the WeWeb team. There are some odd behaviors with the AI at times but Iām sure those will be worked out.
2 Likes
that is great to read! thanks for sharing your experience.
We should ship important improvements later this week, it should get much better, especially with regards to workflows. Weāll keep the community posted!
1 Like
I tried the new AI once in beta, and a few times now after launch, and I didnāt get a usable result even once. So unless there is huge improvements coming, this is a major step back from the āoldā ai that worked really well for making javascript in a workflow, almost always with success.
1 Like
You are right in how getting Formulas often do not work or the AI puts them somewhere else. For instance when Iāve told the AI to use the Conditional Display, it puts the formula in the Text Display rather than the Conditional Display.
We just released a few improvements, especially on the logic generation side. Now it should be easier to use it to create formulas & workflows as well. It is for sure not perfect yet, but we are improving it fast. Let us know what you think of the updated version.
1 Like
Hey @matt-thrive, the updated version should fix this, see @Matthew_Sās video here: https://www.youtube.com/watch?v=_jcA15zDZDY
Let us know if that helped.
For custom components - since our Github accounts can be connected to WeWeb - It would be very cool if we reference a private Github repo for custom component creation with the WeWeb co-pilot. This would allow us to build in ANY co-pilot (Bolt, Lovable, Curson) as a starting point. The use WeWeb co-pilot to bring into WeWeb as a WW no-code component.
2 Likes
could be really nice, but might not be super useful once our component builder is on-par with the other platforms, what do you think?
Weāll be releasing an improved version of our custom component building agent and UX next week, the improvements are substantial, Iāll be curious to get your feedback on this once it is live.
2 Likes