PPQdotAI
PayPerQ

A polished and functional GPT4, Claude3.5, DALL-E (and more) interface that runs on a pay-per-query model via Lightning payments.

Public Key

npub16g4umvwj2pduqc8kt2rv6heq2vhvtulyrsr2a20d4suldwnkl4hquekv4h

NIP-05 Address

payperq@nostr.lol

NIP-57 Address

mattius459@getalby.com

Profile Code

nprofile1qqsdy27dk8f9qk7qvrm94pkdtus9xtk970jpcp4w48k6cw0khfm06mspp4mhxue69uhkummn9ekx7mqpz9mhxue69uhkummnw3e82cfwvdhk6qa04cp

Publishing to

nos.lol

nostrua.com

relay.damus.io

Author Public Key

npub16g4umvwj2pduqc8kt2rv6heq2vhvtulyrsr2a20d4suldwnkl4hquekv4h

Show more details

Last Notes

2025-05-24 14:11:01

There's been some buzz in the last two days around LLM API's runnning pay-per-query via lightning payments.
As the creator of an AI service that prioritizes lightning, I wanted to share my experience and also learn a bit from the audience on this matter.
The ultimate dream we all have in the LN community is for each and every query (inference) to be paid for with the requisite amount of satoshis. That way, the user never has to keep a balance with the service and suffer from the host of corresponding inconveniences that arise from that.
When I originally built PPQ, I tried to implement exactly this feature. But when I actually got to doing this, I realized it was pretty hard:
First, generative AI queries are unpredictable in their cost. When a user sends a request, the cost of that request is generally not known until the output has finished streaming.
Second, even if one decided on some sort of fixed pricing per query, the latency to finish a lightning payment costs precious milliseconds and reduces the snappiness of the end product. I don't want to have a user wait an additional 1 second each time for the payment to clear before getting their answer.
To address this, my best idea was to charge an "extra amount" on the user's first query. That way, my service would store a de facto extra balance on behalf of the user. When the user submits their subsequent queries, the system could draw down on this "micro balance" instantly so that it didn't need to wait for the subsequent payment to clear. This micro balance would also serve to mitigate any issues where the user's output was higher than expected. So each subsequent query would always be drawing down on that micro balance and then the users realtime payments are not paying for the query, they are rathing paying to top up that micro balance over and over again.
However, even this method has some weaknesses to it. How much extra money should that first query be? Theoretically the micro balance needs to be as large as the largest possible cost that a query could be. If it wasn't that size, the service makes itself vulnerable to an attack where the users consistently write queries that exceed the amount of money in their microbalances. But the maximum cost of a gen AI query can actually be pretty large nowadays, esp with certain models. So the user's first query would always have a weird "sticker shock" attached to it where they are paying $1-2 for their first query. It creates confusion.
Aside from these problems, the other big problem is that the lightning consumer ecosystem of wallets and exchanges largely do not yet support streaming payments. The only one that does to my knowledge is @getAlby with their "budgeted payments" function in their browser extension.
So even if you were to build a service that could theoretically accept payments on a per query basis, the rest of the consumer facing ecosystem is not yet equipped to actually stream these payments.
In the end, I just adopted a boring old "top up your account" schema where users can come to the website and deposit chunks of money at a time and then draw down upon that balance slowly over time. While boring, it works just fine for now.
I woud like to hear from the community on this issue. Am I missing something? Is there a better way to tackle this? Maybe ecash has a cool solution to this?
@nprofile…h36r @nprofile…zj6n @nprofile…yvfs

2025-05-22 17:48:30

Boom! Launch of the year!??!
We've already got it live now on PPQ.AI!
https://x.com/AnthropicAI/status/1925591505332576377

2025-05-22 16:39:53

Denominate in Bitcoin.
Spend your Bitcoin.
Pollinate the world with Bitcoin.
https://blossom.primal.net/b150754d5ca87d1ead3150b8f1bce223dcb07385e507c65541c96a521f6d9ab7.png

2025-05-19 14:00:18

Vastly improved our document upload feature. It now supports up to 10 documents and images of various types. We now also allow users to provide links to sources they wish to include in their conversations as well!
https://blossom.primal.net/d43fe75f13f8ff1d6264bcd55fec0ed958fb0f20ab652f016cd66c8251100396.png

2025-05-14 12:13:09

We are looking for engineers and a CTO. Please consider joining!
https://ppq.ai/careers

2025-05-03 17:12:25

- reply

On PPQ the costs of these models are in the same ballpark so I wouldn't normally expect to see this kind of difference.

2025-05-03 17:11:40

- reply

I'm curious what caused this. I know that gpt4o has an automatic caching discount but maybe 4.1 doesn't. Or it could be how goose is somehow using the models differently.

2025-05-01 12:55:41

Did you know that openAI's viral new image model is now available on PPQ?
No credit cards!
No subscriptions!
Pay per use!
Private!
https://blossom.primal.net/16b2ef778f5e66c0aca3c28c7844d236d3f0e8a8fe57dc54df1386ea4e0f5471.png

2025-04-17 09:50:16

- reply

Actually, what happened was that I changed my lightning address in my alby hub to something different but didn't update it on primal. Now it is updated and should work.

2025-04-14 08:36:28

- reply

Our doc upload is pretty hit or miss, sorry about that. We have a plan to make improvements but haven't gotten to it yet. In the meantime, if you simply copy the entire document and paste it into the input box it should do the trick.

2025-04-12 06:10:01

- reply

Yes, what’s up?

2025-04-11 14:56:57

Is there a way to post a note that simultaneously zaps someone?

2025-04-11 11:08:17

- reply

That was quite a turnaround, great job!
What can we do to cross that final barrier of having the invoice automatically load into the nwc wallet and have the LLM send the money along?
I also found these repos:
https://github.com/getalby/nwc-mcp-server
https://github.com/getAlby/lightning-tools-mcp-server
https://github.com/AbdelStark/nostr-mcp
I think you are really close to delivering some great functionality to this MCP server... Some other functionalities that I think would open up a ton of creative use cases would be:
1) Follow a group of users that already follow an account
2) Zap one or a group of users
3) DM one or a group of users
4) Detect specific engagement types on a specific nostr note
5) Keyword search a bunch of nostr notes and zap them accordingly
6) Publish data to nostr in the form of a note
7) unfollow one or a group of users
8 ) like certain notes fitting certain keywords
9 ) send payment to a lightning address
If you can create an MCP that can do these things and more... it could start becoming really powerful.

2025-04-08 16:48:19

Nice work with your nostr MCP server @nprofile…ggfm .
Enjoyed playing around with it so far.
If you add in send zap functionality, we will send you a nice zap using it!
Anyone know of a nice open source repo for an MCP client that we could bring to PPQ?

2025-04-01 09:08:25

- reply

I just DM'ed you @nprofile…2ln2

2025-03-31 07:30:20

- reply

We accept altcoins so that we can build a better Bitcoin product. As someone who has worked with the largest crypto ecommerce company for the last four years (Bitrefill), I know very well the landscape of payments in Bitcoin and crypto.
Why not work with us to build out this great product so that you can get back to working on the other great projects you have?
Much respect!

2025-03-30 21:02:06

- reply

Want us to build it so you can keep focusing on other stuff? Be happy to chat.

2025-03-29 21:21:40

- reply

Should be a couple weeks... when openAI gives us access to it.

2025-03-26 21:31:19

- reply

We will try to get this capability live tomorrow.

2025-03-25 16:20:09

- reply

Thanks for the warm comment!
I've been asked the ecash question before but don't really get it, perhaps because I'm still an ecash newby. hy would I accept ecash when I could just accept lightning? Don't ecash payments convert to lightning? I'm not losing customers by NOT accepting ecash directly, right?
Also, running an mint means running an MSB, right?
Or perhaps you mean that I should run a mint of PPQ ecash tokens or something?
Please explain further!

2025-03-23 16:16:35

- reply

https://primal.net/e/note14luewvy38acq90l46yvtzwzk6y93tly78ullhqpq4llmm86xse5ss6m65g

2025-03-22 08:04:23

- reply

Deepseek-chat is another idea

2025-03-22 08:04:12

- reply

I wonder if gpt4o mini could get the job done correctly? Try that.

2025-03-20 12:46:25

Excited to launch our new Deep Research tool!
Deep Research combines web scraping via firecrawl and iterative reasoning via OpenAI to produce complex reports with citations!
Access Premium AI models and tools via crypto on a pay-per-use basis!
https://ppq.ai/blog/introducing-deep-research

2025-03-13 13:03:22

- reply

Can you dm me the email you used to create the account? Are DM's on nostr private? 😂

2025-03-13 12:58:01

- reply

Thank you for reporting. We are looking into this.

2025-03-11 12:52:31

Hey all, we've created an issue tracker for anyone interested!
https://github.com/PayPerQ/PayPerQ-Issue-Tracker

2025-03-01 19:34:03

Is there a polling function on nostr where user zaps amounts count as votes?

2025-02-28 12:24:32

GPT4.5 is available on PPQ.AI! Specializes in writing and creative tasks.
Careful with this one, it is 30x more expensive than gpt4o! 🤯🤯🤯
https://x.com/OpenAI/status/1895219591070261266

2025-02-26 10:02:32

- reply

Thank you for this great feedback. NWC is on the roadmap for sure. Storing chats in relays I will have to think about. We are more likely to starting using opensecret soon: https://github.com/OpenSecretCloud/OpenSecret-SDK
Custom GPT's on the way
Voice on the way
Auto model selector is on the way but probably not until they good good enough.
Open source I need to be more convinced. Some parts of the code I would be willing to open source for increased transparency, but we need to make money with some proprietary part of our app somehow. ;)
Unsure what scheduled tasks you were thinking of. Can you give me a more concrete example?

2025-02-25 10:56:53

- reply

I just tried it and it seems to be up and working. Can you DM me a screenshot of what you're seeing? Or try again?

2025-02-25 09:30:20

- reply

We highly recommend using Cline in VS Code. You can simply plug in your PPQ API key there. https://m.primal.net/PKbG.png

2025-02-25 08:47:59

Big launch today with Claude 3.7! Likely the world's most intelligent model!
Both normal and extended thinking mode are available on PPQ.AI!
https://www.anthropic.com/news/claude-3-7-sonnet https://m.primal.net/PKZS.png

2025-02-17 16:17:11

Google's Imagen 3 text to image generator is now live on PPQ.AI. It is ranking as the new "best model" on lmarena_ai:
https://m.primal.net/PAZR.png

2025-01-31 12:03:38

Which to choose: @nprofile…cnv5 or @nprofile…fu08 ?

2025-01-28 10:21:26

- reply

Give us a try at PPQ.AI. We have a super slick UI, no subscriptions, pay-per-use via LN, private (convos stored locally also), access to all of the best models in the world, both open source and proprietary, including image and video models.
Also, we do not and will not be creating an altcoin like they have. Our utility token is BITCOIN, our treasury is BTC, our employees get paid in BITCOIN!

2025-01-27 12:22:08

- reply

It takes a while to think sometimes. Up to 45 seconds or so, depending on what context you fed it. Can you try again and see?

2025-01-27 11:13:13

I don't think people realize enough that this is how Bitcoin may die. Not one sweeping action, but millions of little papercut laws that make it worthless to hold.
If the Bitcoin P2P economy doesn't take hold soon, the only on and offramps will be exchanges controlled and surveilled by big gov.
Build faster everyone!
#nevent1q…uxfr

2025-01-23 14:38:02

- reply

https://m.primal.net/NwSU.png

2025-01-23 14:35:48

- reply

You keep mentioning multi modal online for o1, but I don't even believe ChatGPT has that, does it?

2025-01-23 13:57:31

- reply

o1 now available.

2025-01-23 12:51:13

- reply

You guys are talking about something I've been thinking heavily on for a bit now.
First, I'm a HUGE fan of notebookLM... and there IS an open source version of it! Podcastfy.ai!
I have actually thought deeply about how we could bring podcastfy to PPQ.AI but we haven't been able to get to it yet. I even went so far as to have a conversation with a friend about it and he gave me some thoughts on deploying it:
https://m.primal.net/NwLC.png
If either of you are interested, if I created a bounty for you to deploy this for me it would be amazing. If you are interested, let's set a call at some point because there is a lot more to explain.

2025-01-21 13:28:39

HUGE buzz on this new open-source + affordable advanced reasoning model. Now live on PPQ.AI
https://x.com/deepseek_ai/status/1881318130334814301

2025-01-17 08:15:23

- reply

We have 4o and Sonnet already. The versions you are using on ChatGPT are "multimodal", meaning that they can "call" the internet when you ask for it. We added that a while back but then regressed due to bugs. But we could probably try again soon as the tech has advanced a bit in this area.
Additionally, we have o1-preview but are still waiting for o1 and o1 pro to be made available to use via the openAI API. They have been promising that since before christmas but still nothing.

2025-01-15 15:59:18

@nprofile…pfty 's search continues to be bad for me. Been like this for a couple months since their big upgrade. When I search for the term "PPQ" it shows me old results, whereas when I go to @nprofile…7zc2 it shows most recent mentions.

2025-01-07 14:00:44

- reply

A lot of Bitcoiners could use this advice!

2024-12-19 07:07:41

- reply

Do you want the ability for lightning topups, or do you want the ability for actual per query lightning payments? I would think the latter would have an undesirable latency for each query, right?

2024-12-16 12:50:39

- reply

Give PPQ.AI a try. We offer all of the latest and greats AI models, including video models from the best companies, on a pay per use basis via lightning, monero, etc.

2024-11-27 14:59:44

- reply

I need to pay money to use keyword search now? Am I understanding correctly?

2024-11-16 16:42:51

- reply

Yea I know it's free thats why I mentioned "whether via Alby cloud service or your own node". Running own node is also an expense.
I will continue to use Alby irregardless but was just giving my thoughts on the 98% of non-technical, less fee tolerant audience.

2024-11-16 16:10:07

- reply

I love Alby and I think their browser extension in the single most awesome and useful Lightning thing ever created. Having said that, I do think having to pay to be non-custodial (whether via Alby cloud service or your own node), is a total non-starter for 95% of users, and is of course a large reason why lightning has struggled in general.
Additionally, the thing that lightning is best at currently is SMALL payments, which honestly most people don't need non-custodial when it is for small amounts.
Non-custodial is for my cold wallet. I don't mind having some sums in custodial things, so long as they are very convenient and cheap to use.
I'd have happily paid a 500-1000 sats every month to use alby custodially, and I hope they considered that, but guessing it's not possible from a legal standpoint in many jurisdictions.
@npub1get…0nfm

2024-11-13 15:51:15

BOOM! A new era for PPQ.AI as we intro our first line of advanced media models.
These first are world class "text to video" models, and are capable of some WILD generations:
LumaLabsAI DreamMachine
HaiperGenAI Video 2.0
genmoai Mochi 1
Hailuo_AI MiniMax
https://m.primal.net/MCCG.mp4

2024-11-09 15:28:14

- reply

I've got a boy on the way! Looking forward to these times!

2024-11-09 04:40:09

- reply

Our average user spends about $4 per month, but your mileage may vary.

2024-11-08 16:49:30

- reply

Nice, same. I get errors occasionally if the conversation gets too long. Do you as well?

2024-11-08 15:52:47

- reply

Which extension in VS code do you use?

2024-11-08 15:21:35

@PPQdotAI has now solved 50,000 little problems for humans in over 90 countries.
We are democratizing access to premium AI by using ultra-low barrier of entry LN/crypto payments.
Come to us, buy $1 in credit in 10 seconds, and learn from the greatest encyclopedias ever known.
https://m.primal.net/Lwnb.png

2024-10-25 13:22:12

- reply

So how do I go about getting the popular relays to host my users' stuff? I should need to bribe them somehow?

2024-10-25 13:18:34

- reply

I kinda assumed that once they were on one relay, other more popular relays would pick them up automatically. Is that not going to be the case?

2024-10-25 12:58:17

Hey all, I am brainstorming a big idea that involves nostr/ecash but unable to find someone who will answer my dumb questions about it and give feedback.
If you are willing, please like and RT this so it gets some attention. Here is the idea:
Currently, PPQ.AI allows users to do some basic image generation, but it is nothing compared to some of the newer developments coming to the arena soon.
https://m.primal.net/LhlV.png
I am considering creating a whole new side of the webiste, studio.ppq.ai, which will be a place where one can do all sorts more advanced media generations. image, video, 3d, song generation, all sorts of configurable settings/flows, etc. Something like this:
https://m.primal.net/LhkX.png
Where does nostr come in? Here is the idea:
All images generated on PPQ.AI could (optionally) be automatically posted to a nostr relay. How? We (PPQ.AI) will automatically generate a pub/priv key for the user on the client side and have it publish to nostr automatically in their using those keys. We can also generate an ecash wallet for them as well client side (maybe using something like npub.cash). The posts will probably go to a relay that I control (because other relays might see it as spam?), but I think other relays might pick them up as well? The media gens published to nostr will then be viewable to non-PPQ users and also eligible for likes, zaps, etc. This will both encourage my non-nostr users at PPQ to check out nostr for the first time as they get notified of the reactions on nostr to their generations, but it will also encourage nostr users to come check out PPQ when they see AI generations which leveraged it. I expect 90%+ of PPQ users may never actually get onto nostr and check out the reactions to their images, but some will, especially if they get zapped a few sats for a cool generation, and that may lead to furthering nostr.
In addition to posting to nostr, I would also create a "community browsing website" where one can all of the trending images, like them, zap them, utilize their prompts, etc. It will look like this web page (https://www.midjourney.com/showcase), except when the users click the image the nostr note metadata for that image will show up in the sidebar.
https://m.primal.net/LhkT.png
Please let me know what you think of this idea as well as the questions about relays and ecash:
- If I push user-generated media to a relay, it may reject them as spam, right? So do I need to do my own relay or perhaps I need to pay a relay to host my users' generations?
- I can create ecash wallets automatically on the client side, right? Which library might I use for that?
- If I run the mint, does that put me in any sort of legal peril as custodian? If so, are there random mints I can pawn this off on to avoid legal peril?
- How do I connect to an API that a relay provides? Where is the best guide for that? I found @npub1sx9…6g8m but the endpoints it has do not provide me with the types of stuff that I'm looking for.
Thank you for reading, please give some feedback on the idea, and please like/RT for visibility.

2024-10-23 23:50:58

- reply

Jan.ai with a PPQ API key. 😉

2024-10-10 19:24:19

- reply

Have we seen rugs yet with ecash?

2024-10-10 19:21:30

- reply

An Artifacts-like addition is something we've looked at, but honestly I can't find strong use cases for it being making snake games. Can you give some examples of the times that artifacts has been a really handy tool to have?
Suno doesn't have an API yet, but other music gens I've looked at.
I've also seen Leonardo... we were considering building a studio.ppq.ai side where it has a more extensive media suite.
Just a matter of priorities!

2024-10-10 17:03:10

We now accept Monero (XMR) via our @BtcpayServer integration!
Use #Monero to access all of your favorite premium AI chat and image models.
- Pay-per-use
- No subscriptions or credit cards
- Data is stored locally and not on our servers!
https://m.primal.net/LRSq.png

2024-09-17 14:31:42

Ideogram AI is now live on PPQ.AI! Ideogram is a premium image gen model and is especially good at spelling out words, something that other image gens struggle with.
The meme potential is high with this one!
https://m.primal.net/KsoK.jpg

2024-08-30 19:13:04

- reply

Unfortunately it doesn't give access to the premium models or any media generation. But it's a nice freemium bot to use.

2024-08-30 19:09:01

- reply

Never heard of this. Will look into it!

2024-08-30 19:07:07

Venice.AI has cut off freemium access to their Llama 405B. For acess to 405B, gpt4o, Claude, and others, come to PPQ.AI.
No accounts, subscriptions, or credit cards.
Private
Pay per use
On Lightning ⚡
https://m.primal.net/KVkF.png

2024-08-30 12:54:00

- reply

Give us a try! Access to all of the best models and more but without a pesky account, subscription or credit card connection.
DM me if you'd like some free credit to get started.https://m.primal.net/KVJG.png

2024-08-30 12:50:40

- reply

Give us a try! Access to all of your favorite models and more but without a pesky subscription and credit card connection.
DM me if you'd like some free credit to get started.
https://m.primal.net/KVIt.png

2024-08-29 15:17:56

- reply

FWIW I got more followers and engagement from this tweet than all the zapvertising / satograms I'd ever done before. Your idea of giving substantive money doesn't work as well because it gives people pure money with no strings attached instead of credit on my platform. I'd rather give them credit since it "dog foods" my app and I only lose it if they actually redeem the credit, unlike zapvertisements. Additionally, I have a bunch of free credits from cloud platform anyways to burn through so again giving away credit is the better option in this use case.
There are many more ideas I plan to employ using zaps and such, but just wanted to give you some real feedback on how it works in practice.

2024-08-28 13:34:38

- reply

Hey Derek, I previously tried using @npub1g24…pfjl with kind of marginal results. The guy suggested that maybe my account wasn't "warmed up" enough at that time.
Still, Zapadd only allows for me to zap a random group of popular tweets, but what I really wanted was to Zap people who specifically mentioned AI, or who follow other AI nostr accounts. No project thus far has built out a platform for doing that.
In addition to Zapadd, I've also zapped dozens of folks who mention AI in some form, but that is more of a manual process. I'd rather use an automated way!
Lastly, I've tried used satograms before as well!
https://x.com/MattAhlborg/status/1775527923711111621

2024-08-28 13:04:47

Zaps on Nostr can create some really interesting marketing opportunities for companies trying to gain users. Let's try an experiment:
For every user who zaps or reposts this, I will DM you $1 in credit for use on PPQ.AI.
https://m.primal.net/KReX.png

2024-08-28 12:43:49

- reply

PPQ.AI is an accountless, subscriptionless, pay-per-use AI Chatbot which runs on Lightning!
You can use GPT4o, Claude3.5, DALL-E, as well as a host of more open models like Llama, Mixtral, etc.
All of your data is stored on the client side!
Please give us a shoutout!

2024-08-27 20:33:34

Is there a plan to add "delete note" functionality to nostr? None of the clients I use seem to support it. I get that ultimately it is the relay's decision... but still seems like could be a thing to add.

2024-08-08 18:56:23

- reply

You can use the BEST models in the world (Claude, GPT4o) on PPQ.AI for pennies via Lightning.
No accounts, no subscriptions, pay-per-use.

2024-07-24 17:40:50

Sexy new carousel:
https://m.primal.net/Jcwa.png

2024-07-17 15:45:20

- reply

Follow me!

2024-07-13 13:47:47

https://pollerama.fun/respond/688c0a279d6424a5a610c51ca0e4ec24cb407eaa01f42f95977fbabe065591bb

2024-07-12 10:51:07

1/ We now have INPAINTING! INPAINTING allows you to use AI to edit images in fun and interesting ways.
Use the inpainting tool on PPQ.AI to edit any image, real or AI generated
In the example below, I've adjusted for the poor lightning to show my true arm size!
https://m.primal.net/JNwK.png

2024-04-30 12:31:22

https://stacker.news/items/524200

2024-04-24 20:29:39

- reply

No login required at PPQ.AI.

2024-04-24 17:54:44

PayPerQ (PPQ.AI) is a polished and functional AI Chatbot that runs on a pay-per-query model via Lightning payments. It supports both closed and open sourced models, image generation, and many of the other amenities you would come to expect form a premium AI experience.
https://image.nostr.build/8bb9429bfb087d30c1b3e98c4f22e1a8465aaaa8b7033a2df867469fa52d6f68.png

2024-04-24 17:04:59

- reply

We're on nostr now!

2024-04-24 17:04:24

- reply

We're on nostr now!

2024-04-24 17:04:13

- reply

We're on Nostr now!

2024-04-24 17:03:51

- reply

Hello!

2024-04-24 17:03:30

- reply

2024-04-23 20:52:46

Test. Can anyone hear me?