Love the idea. Simple and beautiful.

As GPT is used in all text-related activities to give feedback to the writer, GPT-4V in the future will be applied to all websites, posters, book covers, graphs, plots, design of hardware devices, movies, computer games, etc.

I suspect that teams at Figma, Canva, Adobe, and others are already working on it.

It looks like GPT-4V is not just another model but a change in infrastructure, unlocking new possibilities and new "blue oceans." as any good new multi-modality.

It feels so great to live in the future.

Yeah, I think there are limitless possibilities. Yesterday I saw on twitter someone using GPT-4V to narrate a video game in realtime, just mind-blowing. The future is here.

Very cool idea! One question, how much is this costing you out of pocket to host this? The OpenAI GPT4 part must be not nothing.

Thank you!

Currently it's about:

* Rendering the screenshot (urlbox): $0.02

* GPT-4V 1K input tokens = $0.01

* GPT-4V 1K output tokens = $0.03

* GPT-4V Vision pricing (per image size): $0.008

Total: 6.8 cents per a single audit

May not seem like much, but I see a lot of traffic coming from HN, so say 10K people use it, it'll cost $680...

Eventually, Hacker News ended up bringing 20,000+ visitors. Had to upgrade our urlbox screenshot API account twice in 1 hour just to keep up with the demand.

This Hacker News post could have cost $1,000+ if the API rate limit hadn't kicked in. Oh well, silver linings...

Since almost no one could use the app (too many people, and not enough quota from OpenAI), I added a $1.99 one-time payment per audit. This would probably decrease the number of users and also help cover the insanely high costs.

Thanks for following up, I've had ideas for something in the same light, but need to be careful not going bankrupt!

Good stuff! I actually had the same idea as I was building a custom GPT. Not as slick as your effort but if you want to take a look: https://chat.openai.com/g/g-lwQixzKwd-web-optimizer

Oh, this is pretty cool! I was hoping to try and create a custom GPT too, but couldn't resist trying the API before.

I tried to use this and it showed me an apache2 landing page instead of my app website, which led me to realise my app site had been down for at least a day because apache had actually crashed due to nginx using port 80 (I don't even use nginx so not sure what the heck happened)

Anyway, not the success story I'd hoped to post here, but it definitely solved at least one problem I didn't know I had.

I'm now going to add uptime monitoring

Hey Rory, are we talking about the same app? https://uxaudit.vercel.app

It's really weird that you've seen an Apache2 landing page, as this page is hosted on Vercel.

Do you still get an error message while visiting the same page?

I'm getting a 500 back from /api/audit, so I'm assuming that it's already more popular than expoected :).

Yes, just getting a lot of traffic from HN. Try again tomorrow, please? Error: 429 Rate limit reached for gpt-4-vision-preview in organization org-..

Sad to hear that. Probably this will be a common problem: good ideas probably dying/losing a lot initial traction because of being rate limited on launch.

Yep, this is my life now: https://twitter.com/LiorGrossman/status/1722681560128762208

I think you only get 100 requests / day during the preview.

Yep, it looks like it :( Any idea how to overcome this limitation?

OpenAI recently updated their limit documentation so there are tiers now. You automatically scale up through the tiers but you may be able to jump the line a bit by paying some extra. But I think you will have to wait 7-14 days since first payment before going to next tier. Try finding the usage and limits docs in your account.

Edit: Oh I see I'm in Tier 3 and the RPD limit on vision preview is still 100/day. Guess that's just it for now. It is labelled preview after all.

Thanks. Yeah, it seems that even on Tier 5 the limit is 100/day, that's not enough for any production app. I hope they go out of preview mode soon and therefore increase the limits

Boy, that escalated quickly...

Just deployed a version that shows some UI indications of OpenAI limit errors.

Neat. I think Aaron Levie (CEO of Box) pointed out this capability of GPT-4 on Oct 13: https://twitter.com/levie/status/1712882032290807835

Yeah, I feel like the Vision API will unlock many new opportunities. Honestly, it's been super easy to work with it. Just sad that they have such a low rate limit in the preview phase.

Getting a 500 error - API limit?

Yes, just getting a lot of traffic from HN. Please keep window open and try again in 15min, sorry...

Error: 429 Rate limit reached for gpt-4-vision-preview in organization org-... on requests per day (RPD): Limit 100, Used 100, Requested 1. Please try again in 14m24s. Visit https://platform.openai.com/account/rate-limits to learn more.

I get an HTTP 500 on Chrome. On Firefox, I just get a NextJS NetworkError when it tries to fetch the screenshot. I'll try again later with Chrome

Thanks, it's OpenAI's rate limiting indeed.

Just deployed a version that shows some UI indications of OpenAI limit errors.

Please try again later today.

"Please let me know your thoughts and suggestions on how to improve it"

I found this cool tool that offers advice on webpages. Here's what it said.

Suggestions

Problem:The page has a lot of unused white space which can make the content seem sparse Suggestion:Utilize the space more effectively with visuals or additional information about the service DESIGN USABILITY

Problem:The headline "Instant UX audit for your landing page" is generic and doesn't convey a unique value proposition Suggestion:Refine the headline to clearly articulate what makes this service unique or better than competitors MESSAGING BRANDING

Problem:The subheading "Get actionable suggestions to improve your usability, conversion and messaging" is informative but lacks impact Suggestion:Make the subheading more compelling by including specific benefits or outcomes users can expect USABILITY MESSAGING

Problem:The page lacks content that could be indexed by search engines for better discoverability Suggestion:Add more descriptive text, headings, and keywords relevant to the service for SEO purposes USABILITY SEO

Problem:There is no clear indication of what happens after entering a URL Suggestion:Provide a brief description or visual cue about the process that follows the URL submission USABILITY CRO

Problem:The input field for the URL lacks a label, which may confuse users Suggestion:Add a clear label above the input field to indicate where users should enter their URL USABILITY CRO

Problem:The color contrast between the background and the input field is low, which may affect visibility for some users Suggestion:Increase the contrast to make the input field more prominent DESIGN USABILITY

Problem:There are no alternative text descriptions for interactive elements, which is important for screen readers Suggestion:Ensure all interactive elements have descriptive alt text for accessibility USABILITY ACCESSIBILITY

Problem:The call-to-action button "Start UX Audit" is not immediately visible without scrolling Suggestion:Move the call-to-action button higher up on the page to ensure it's seen without scrolling USABILITY CRO

Problem:The page does not provide any social proof or testimonials which can help build trust Suggestion:Add customer testimonials or logos of companies that have used the service

Yo dawg, I heard you have a UX audit app, so I ran the UX audit app on the UX audit app, so it does a UX audit to the app!

seriously I would add some kind of screen shots in a carousel below just to give people more of an idea of what it does.

Yeah that's a good point. I wanted to keep the UI clean (Google search alike), but people don't even know what the results look like.

Also, add an email me the results below the results. or save to my account.