Top
Best
New

Posted by rvz 17 hours ago

OpenRouter is joining Stripe(openrouter.ai)
Previously: Stripe will reportedly acquire OpenRouter for $7B+ https://news.ycombinator.com/item?id=49323381
858 points | 444 comments
nikcub 12 hours ago|
I love OpenRouter, long time user. Stripe will hopefully be a good custodian.

I just want to point out some features of OpenRouter that make it more than just a model selection and routing endpoint and that I find incredibly useful:

0/ Default routing is to the cheapest provider, but they're usually not the most performant. I'd guess 99% of OpenRouter integrations never tweak the default routing. Here you can setup cheapest with performance minimums:

https://openrouter.ai/docs/guides/routing/provider-selection...

You can also stack model selection in priority

1/ Using broadcast you can push all your analytics to clickhouse / s3 / snowflake and a bunch of other compatible destinations. Setup a clickhouse server ($5 VPS[0]) and send all your traces to it:

https://openrouter.ai/docs/guides/features/broadcast

customise your own observability in your dashboards from there. Superwin

2/ Model router is also a natural home for llm security - OpenRouter has the beginnings of prompt injection detection:

https://openrouter.ai/docs/guides/features/guardrails/prompt...

there is also PII detection. This will show up in observability as rejections/blocks etc.

There are so many model routing solutions (same with observability, security etc.) but they're all 80% solutions - OpenRouter really rounds out with well implemented features that you need when deploying models at any scale and I gladly pay the toll.

[0] not sure if these exist any more but clickhouse is resource efficient

hobofan 4 hours ago||
> Model router is also a natural home for llm security - OpenRouter has the beginnings of prompt injection detection

Maybe I'm biased from the perspective of a "harness provider", but I think Model routers often have too little context to act as well-informed prompt injection prevention. e.g. it lacks context of where which part of the message(s) originates from and sanitization/safeguards were already performed on the application layer.

Something like OpenRouter's "flag" mode is fine, but usage of auto-redact or auto-block should really only be used if there is significant risk exposure through your harness or otherwise they are a constant source of bugs.

dayeye2006 6 hours ago|||
noob q -- does the routing hurt cache hit rate?
Ey7NFZ3P0nzAe 4 hours ago|||
By default yes bit you can pin providers to avoid this.
miroljub 5 hours ago|||
Of course, of routing the same session to different providers.

But don't worry, many providers there charge for cache the same price like non cached ;)

sbinnee 6 hours ago|||
I have been using openrouter only superficially. I didn’t know about the broacasting feature. Thanks for that.
johnbarron 3 hours ago|||
They might be popular for indie developers, but nobody doing serious AI or in a corporate environment is using them.

And if they are, their compliance team is about to strike them down. The VCs forcing this acquisition do know this.

- Why would you add a penalty of 50 ms at a minimum? And that is not the p95... Just run LiteLLM in house and you dont even really need that.

- Their capacity pools are shared across the whole user base, a massive batch processing by another of their customers and think what that means for your response time...

- So instead of negotiating corporate rates with OpenAI or Anthropic, you would be using an intermediary and topping up the corporate credit card... for a 5% markdown ? Really?

- They can see all your critical corporate data on the in and out

- They present some pink SOC 2 promises but then wash their hands and defer to you and the providers. Its just the Bolt and Uber model the drivers are not our employees....

- They are a man in the middle proxy that is a massive security liability for your corporation

- They have no support for private cloud points

- No geofencing guarantees

- No intellectual property legal indemnification unlike what AWS or Microsoft or Google offers

- Its a provider roulette inconsistent with hosts providing different quantization levels causing random shifts in response quality

- Support via a Discord server...

The only reason they were not shutdown yet by Anthropic or OpenAI is because they have the same VCs, as those two. That would mean said VCs investment would go to zero. Oh and those are the same VCs that own Stripe...

Just setup a private proxy tier using something like LiteLLM, even if you really dont need it. Just code your enterprise apps to have have fallback loops on the core hyperscaler providers like AWS Bedrock or Azure Foundry...

devy 2 hours ago||
Yep. https://news.ycombinator.com/item?id=49344067
troupo 5 hours ago|||
> features of OpenRouter that make it more than just a model selection and routing endpoint

All those features show it's just a model selector and router.

johnbarron 3 hours ago||
Meaning you can switch in seconds. Stripe just has too much money...
pbreit 5 hours ago||
This seems like something xAi could reproduce in a weekend
rswail 4 hours ago|||
Could but hasn't. What exactly is xAI's USP compared to the others?
michaelt 4 hours ago||||
The other thing OpenRouter gets you is: Access to loads of models via a common API/sign-up/prepayment mechanism.

I want to know which vendor/model does best at my extracting-facts-from-text task? Which does best at my OCR-a-text-document task? Which can deal with a safe-for-work beach photo without a censorship system false alarm?

OpenRouter lets me run my tests against openai and anthropic and google and x and bytedance and qwen and llama, with a single sign-up and a single payment.

catlifeonmars 2 hours ago||
You can do that with AWS too.
mcintyre1994 3 hours ago||||
Nobody would trust one of the labs that have their own models with this. They’d assume they’re taking all the data, and biasing the routing to favour themselves. xAI would probably be the least trusted, but the same would apply for any of them.
Flamkuchlo 1 hour ago||||
Everyone could reproduce this on a weekend...
ahoka 1 hour ago|||
xAI just not in the game, never was.
freely0085 5 hours ago||
Using Open in your company name, when there's nothing about it, should be illegal.
razemio 2 hours ago||
I think it is fitting because they provide „all“ models independent from the provider. Open in this sense means to me being open to switch the model, speed and provider at any time.
fungi 5 hours ago|||
And don't get me started on Sesame!
open-sesame 3 hours ago||
Me?
tsoukase 4 hours ago||
I used to demand the same but a word can have multiple meanings depending on the context, eg OpenMind means something specific. Open{$IT-term} does not necessarily mean Open Source {$IT-term} and OpenAI might mean an AI available to all people, not implying anything about cost. What could be illegal is the word Free, which a for-profit company would never use in the first place.
Ey7NFZ3P0nzAe 4 hours ago||
Found a good european alternative: https://cortecs.ai/pricing
danvdb 1 hour ago||
Probably the main EU alternative to Openrouter is actually https://requesty.ai, more fully featured and equivalent model access.
crossroadsguy 41 minutes ago|||
Indeed. The previous one doesn't even have glm 5.3 listed.
nemoniac 1 hour ago|||
It looks like it's based in London, not EU.
barrell 2 hours ago|||
Awesome will check this out. EUrouter is too cost prohibitive for me, and now that open router has been acquired by stripe it’s only a matter of time until it’s unusable
mbi 1 hour ago|||
They seem to be about twice as expensive on a few openweight models I've checked.
dethos 3 hours ago||
Thanks for the suggestion.
apexalpha 15 hours ago||
Great product, been using it for a while.

Turns out even a proxy can be worth $8bn with the right business model behind it.

Users get an array of providers competing behind a single API, meaning they have to compete on price and quality not vendor lock-in. This encourages users to join OpenRouter over specific model vendors.

Providers get easy access to revenue (and data) and new customers with little to no ad spending, encouraging them onto the platform too.

And that's all you need. Win win.

Well done and congratulations.

giancarlostoro 11 hours ago||
Someone noted they pass through 100 billion annually if I remember correctly, but they also take a 5.5% cut, so Stripe secures them as a “customer” while getting their money back over time. It also to me feels like an adjacent business aligned with sort of what Stripe already does to begin with, being someone elses middle man.
throw03172019 8 hours ago|||
That feels high if they were bought at $7B? 5.5% of $100B is $5.5B in annual revenue. 1.3x multiple for AI. No way
byzantinegene 7 hours ago||
100b sounds implausible if you consider that anthropic's revenue is only expected to reach 100b this year, 100m is more likely.
alex43578 4 hours ago||
100b is too high, but 100m is way too low.
johnbarron 3 hours ago|||
Who is that someone? And unnamed employee that told somebody else?

They make less than 140 million in revenue, what today is like 5 signing bonus for OpenAI :-) They are not profitable and see through less than 2 billion in revenue.

These are easy to check so...

michaelbuckbee 13 hours ago|||
I've been using OpenRouter since relatively early (I build evvl.ai - an eval platform on top of it) and think that they're selling at a good time.

An underpinning of their model is that API calls / inference are similar across providers allowing for commodity tokenization cost comparisons, but the providers are beginning to shift to non commodity features that don't easily shift.

For instance, calling Gemini with "search grounding" isn't something that OpenRouter can do (they sub their own web search in), last I checked they weren't doing real time voice models, etc.

lxgr 13 hours ago||
> For instance, calling Gemini with "search grounding" isn't something that OpenRouter can do (they sub their own web search in)

It can: https://openrouter.ai/docs/guides/features/server-tools/web-...

It's apparently even the default these days (which makes sense, as it's usually better in my experience).

michaelbuckbee 12 hours ago||
Nice. Happy to be mistaken on that point as that seems very useful.
Whiteshadow12 11 hours ago|||
Your main point is still right, the providers are beginning to shift to non-commodity features.

Another example might be all the managed agent features, where you get a VM + model, right now it's mostly Google and Anthropic that have this offering.

The service tier is another, where Google, Vertex, OAI, Anth offer it but only OAI and Google offer flex service tier and not all models from them get it.

OpenRouter for me is an abstraction over that complexity, they will have their work cut out for them.

mandevil 10 hours ago||
Fortunately Stripe is in the business of providing a single abstraction over an enormous pit of complexity, so I think they'll be in good hands!
0x457 8 hours ago||
I can see Stripe adding extra services on top. However, in their main business its "easy" for them because getting into payment processing is hard intially. But whole search, harness, VMs, agent loops are not hard to find somewhere else.
jatora 11 hours ago|||
[flagged]
Flamkuchlo 1 hour ago|||
I think it has to be more than just a 'proxy' what Stripe bought.

Stripe and every other big company could build this without any issues.

It either is just a really stupid business decision or it is about the name. The only model proxy i know is OpenRouter despite plenty of other model proxies existing.

hoppp 11 hours ago|||
It's a very straightforward business model and they started just in time to become popular as there are hundreds of alternatives out there.

I also congratulate the founders for pulling this off.

rajnathani 5 hours ago|||
> Turns out even a proxy can be worth $8bn with the right business model behind it.

People said a similar thing for Cursor simply being a VSCode wrapper for when it crossed unicorn status.

999900000999 5 hours ago||
Microsoft is really really bad at marketing and making products consumers actually want.

I really think they only know how to sell Azure.

No reason Microsoft couldn’t have shipped a better version of Cursor considering they wrote the IDE cursor is based on.

croes 12 hours ago||
Given the costs of the providers I doubt that sharing the revenue is a win for them.

They are fighting to be the one and only

bdangubic 12 hours ago||
no company will reject revenue :)
croes 12 hours ago||
But every company wants more revenue.

The smaller, less known could get a win, but the bigger lose.

jamest 2 hours ago||
Alex (the OpenRouter founder) and I went for walks to Sightglass coffee on Divisadero St during our weekly office hours @ HF0 in the Fall of '23.

On one walk I asked 'why are you doing another startup?'

For context, his last one, OpenSea, was valued well into the billions, so it wasn't for money.

His reply: "I just love solving all the puzzles."

It's incredibly hard to compete with someone who is playing the game for the love of the game.

Kudos on playing well, Alex.

Flamkuchlo 1 hour ago|
OpenSea is full of scams, has not added any real value to anyone or society.

DOn't get me wrong, great for him to make money and being able to move fast and succeed, but you could play this game to if you want.

namjh 5 hours ago||
OpenRouter is great business I agree, but I'm not still convinced how the integrity of providers' models is ensured. In other words, can't the provider serve DSv4 flash advertising it as DSv4 Pro?

I'm aware that OpenRouter checks response quality onboarding, and does further checks occasionally, but I'm concerned that it's basically a cat-and-a-mouse problem between the scammers and the detectors. For example, there could be a signal that a specific pattern of requests are from OpenRouter's quality testing bots. Or, they can just route 1% of requests to an inferior model and benefit a small gain, hoping it fits into the statistically allowed margin.

pixlmint 5 hours ago||
How is that a problem that's unique to openrouter? Who's to say when I query the Anthropic API directly, wanting to use Opus, but they determine it doesn' need Opus, route it to Sonnet instead and pocket the difference? The only difference with openrouter is that it's another layer where this type of fraud could occur.
ajyoon 4 hours ago||
OpenRouter's list of providers is very large and includes many little known and sketchy companies.
mmoustafa 5 hours ago||
Yes, it is awfully inconsistent today. They run some tests (accuracy table buried on model page) but they are sparse and only capture a single moment in time. I would love to see OpenRouter take this more seriously.
powvans 16 hours ago||
AI products are going to have to deal with accounting. An agent performs some work. It uses some models and perhaps some other metered services. Someone has to meter that activity. They have to attribute costs, apply the product's pricing rules, bill the customer, collect, reconcile with vendors, and maintain the ledger.

Stripe can use OpenRouter to build the financial and accounting infrastructure for every product that sells metered AI work.

I think the analogy is ADP. Payroll for all the work that's going to be done by AI agents.

aleqs 15 hours ago||
Every token/llm provider already has usage metering...so not sure what exactly you're imagining stripe adds here
ygjb 14 hours ago|||
A single policy location for multiple models with usage limits/rate limits, spend limits all configurable using a common syntax + deep integration with account levels and high visibility into subscription/payments related to meter usage to help understand whether or not your pricing model is underwater as costs change is exceptionally valuable. AFAIK each product consuming model providers without an openrouter style proxy has to do that work themselves, and if you are already using stripe for subs/payments, the direct integration is a huge value add for customers, especially if you can alert on "upcoming price changes from your provider changes your unit economics for XYZ packages/sub tiers".
aleqs 14 hours ago||
You're just describing what OpenRouter already does?
ygjb 12 hours ago||
The first half yes, the second half, I don't think so, because that would require OpenRouter to know what you are charging your customers.
selvan 6 hours ago||||
Similar to how Stripe is a middleman for payments across (fragmented) banks, they want to be a middleman for (fragmented) AI models as tokens are the new currency.

As AI agents/harness/human are spenders of tokens, enabling them to derisk from being locked to a specific model provider & allowing to (re)route to any model at anytime for better leverage in a single API, similar to how they are doing for payments.

andruby 4 hours ago||||
paying for data access, for "deterministic computing", for letting the AI use online services. Maybe even micro payments for using tested and verified skills/mcps.

Example: access to real-time stock trading data, access to weather information, letting it make stock trades, etc.

At first I was confused why Stripe bought OpenRouter, but I think this makes sense.

kridsdale1 15 hours ago|||
Aggregation behind one interface, just like the actual token product.
aleqs 15 hours ago||
That's what OpenRouter already does, so the Stripe value add here is still unclear
pests 14 hours ago||
Not OP, but I could see the vision. Right now OpenRouter is developer-paid with no way to pass costs onto the end user. Stripe could help bill and meter at the user level for per-user billing. I would also think more SDKs/easier ways to add AI features (paid for via openrouter) into apps and websites.
aleqs 14 hours ago|||
Open router supports per-request usage metering, as well as per-key. Building user-level metering from this is trivial, and there are many implementations of this out there. Supporting per-user first-class is trivial considering Open router already has per-request and already supports per-request attribution. How important that is, is a big question, but definitely not something that requires Stripe to build.
user43928 13 hours ago||
Could there be use cases where you want to run inference for users in your app, but don't want to host your own backend, user authentication, and billing?

The user could log into his personal OpenRouter account, authorize your application, and optionally set a budget, all in something like a Stripe payment screen.

aleqs 13 hours ago||
> The user could log into his personal OpenRouter account, authorize your application, and optionally set a budget, all in something like a Stripe payment screen.

Like standard oauth?

https://openrouter.ai/docs/guides/overview/auth/oauth

user43928 4 hours ago|||
Good point, just needs an user friendly flow to sign up and set the budget in a way that is more similar to Stripe checkout than regular account creation.
devmor 13 hours ago|||
OpenRouter already supports this, requests return their exact cost.
nevi-me 15 hours ago|||
IDK, that sounds like a solution in search of a problem. What would be different about this than what OpenRouter is currently doing with pricing?
robocat 13 hours ago||
Metering as a service?

Perhaps stripe thinks metering is a lovely big market? A nice complement to their existing service. Plus a little AI buzz likely helps their valuation. OpenRouter could be a step rather than a goal.

CTDOCodebases 4 hours ago|||
It's not just that.

As user agents become more common it's only natural that they will be used for taking the heavy lifting out of e-commerce purchases. There will be a big need for digital payments to verify and reconcile these purchases.

With LLMs that are plugged into digital payments we will essentially have buying agents in our pocket that can find us exactly what we want for the cheapest price and the quickest delivery.

j45 15 hours ago|||
At some point AI will be able to deal with the repetitive and template based parts of accounting better than accounting.

Which is saying a lot.

chairmansteve 14 hours ago||
I'm not sure that's true. LLMs seem too unreliable for accounting.
RealStupidity 9 hours ago||
I'm always surprised at how many people in the tech sector don't think that things can and will change. In the next couple of years? Sure. 5+? Who knows. Necessity is the mother of invention.
DrewADesign 8 hours ago|||
I'm always surprised at how many people in the tech sector assume overcoming fundamental LLM weaknesses is always “just a matter of time,” and that time will arrive quickly enough to be relevant. If you haven’t seen any of the videos of people trying to get frontier LLMs to count consistently, you might think we’re further from LLM accountant than you realize. It’s not a data problem— I’m pretty sure they’ve encountered numeric progressions between 1 ABs l and 100 before.

And there is no necessity. We have accountants.

RealStupidity 7 hours ago|||
I'm not just talking about LLMs though. People thought the internet and email were just a fad that would never catch on. People thought that the idea of home computing was absurd. All because they only saw what was in front of them at the time.

Also, how many accountants do you know that are truly happy with their work? I'd like to think many of them would love to do something other than crunch numbers all day.

oblio 7 hours ago||
> People thought the internet and email were just a fad that would never catch on. People thought that the idea of home computing was absurd.

We're far removed from "we've sold all the computers the world can buy" (back from when computers were the size of a building).

This is a disingenuous argument. The web and email were basically instant hits and people realized it. Similar story for home computing once the computing power caught up.

oblio 7 hours ago|||
> I'm always surprised at how many people in the tech sector assume overcoming fundamental LLM weaknesses is always “just a matter of time,” and that time will arrive quickly enough to be relevant.

Lots of techies are tech-optimists ("tech always improves quickly").

Lots of people also have dollar signs in their eyes (or related, such as increased visibility and scope).

Hard to tell which is which.

seanclayton 9 hours ago|||
Things can easily change for the worse, such as a stock market crash. Who knows? Will we get AGI or will we see more homeless? History says... AGI for sure!
wat10000 8 hours ago||
It's very difficult for this sort of thing to change for the worse. It's basically impossible on the software side. If next year's model is worse then keep using this year's model. Hardware technology can potentially go backwards, but things have to get really bad for that to happen.
oblio 7 hours ago||
LLMs are inherently unreliable. We do already work with unreliable technologies for many things but there's this engineering principle called "use the right tool for the job".

And it's a lot more likely that LLMs can't be changed to become unreliable, it's just how they work. So we would need more basic research, that doesn't grow on trees and for which the timelines are basically open ended. Maybe tomorrow, maybe right after cold fusion hits mass adoption.

shanev 15 hours ago||
I built modelrelay.ai that does this. Happy to give you an invite code if you want to play with it.
marcsnid 16 hours ago||
I'm not happy about this. Having used OpenRouter a lot and enjoying the experience, I need to say that corporate consolidation is never good for consumers.
rnd0 15 hours ago||
>I need to say that corporate consolidation is never good for consumers.

Truer words, never spoken. I'm not sure how exactly this will screw me over -but I do know that it will.

hstaab 11 hours ago||
Then switch. The good news is its’s fairly easy.
rnd0 10 hours ago|||
Great advice, until all of the alternatives are driven from the market ala youtube, ala' ebay, etc etc
ifwinterco 3 hours ago||
EBay there are alternatives (gumtree etc.).

YouTube there’s no real alternative because hosting unlimited amounts of video for free is a money pit that’s impossible to turn into a viable business.

If some big corporation wasn’t willing to subsidise it for some ulterior motive, it simply wouldn’t exist

HDBaseT 11 hours ago|||
What other routers exist that are good in your experience?
Computer0 10 hours ago|||
Eurorouter
Ey7NFZ3P0nzAe 4 hours ago||
Can't find anything by that name
noname120 2 hours ago||
Maybe he meant EUrouter? But it seems to be an entirely different thing: unified API for models vs a token marketplace
Phelinofist 5 hours ago|||
Cortecs.ai
deaux 40 minutes ago||
Only EU providers, seemingly huge latency, >2x pricing is just in no way an alternative.

https://cortecs.ai/detailedServerlessView/deepseek-v4-flash-...

https://openrouter.ai/deepseek/deepseek-v4-flash-0731#provid...

goalieca 16 hours ago|||
curious as a non-user, but what does openrouter have to do with stripe? Ignoring the corpo-speak about mission statement.. the missions don't really seem aligned in reality.
marcsnid 14 hours ago|||
I don't think they have much to do with stripe. I'm guessing they want to round out their business more and include payments for AI since OpenRouter is basically just an AI traffic router that takes your money and provides it to the APIs you use.
tyre 12 hours ago||||
I wrote a substantial comment about this a few days ago: https://news.ycombinator.com/item?id=49324405
abigdog 12 hours ago|||
My guess: your LLM agent makes buying decisions for you. Providers of agents can integrate witg stripe. Those using openrouter click a button for micropayment support.
bluecalm 3 hours ago|||
That's a rule but then Stripe won the market by having very friendly interface and minimizing the pain of dealing with paperwork for merchants. When PayPal (the worst offender, just hostile company) acquired BrainTree it was clearly a terrible news. When Stripe acquires OpenRouter I wouldn't be so sure.
applicative 11 hours ago||
This must a purely aesthetic judgment apriori. Many services of value to me failed for lack of capital.
Oras 15 hours ago||
Happy for the OpenRouter team. Been using the platform since early stages and the ability to run any AI model with the same api key and same request has been great to experiment with new models and being able to switch models in prod with minimal effort.

They also support fallback by default so you don’t have to write wrappers and logic to choose models, it just works with their SDk using config.

Computer0 10 hours ago|
The open router team was good people with a good product. Hopefully they can move on with their lives
hn3ufz62f7 10 minutes ago|
Filing under things I wish I knew earlier
More comments...