The fact that they put "every model" repeatedly, then don't actually support even close to "most mainstream models", makes me highly doubtful their service will be more honest and forthcoming once I'm signed up.
The domain was previously owned by Innovation HQ (innovationhq.com) who own tens of thousands of domains, some of the best in the world (they sold robot.com not long ago, owns/runs whatismyip.com). Innovation were asking ~$5 million for it earlier this year. Given the huge money flowing into AI router projects, I expect Ramp paid very close to asking, or perhaps a little more. So, $5 million +/- $1 million is my guess. I assume Ramp will announce the price in a few weeks to generate some more press for the project.
I think you mean landlords, not squatters. Squatters are the people who occupy owned property without permission, landlords are the people who hold property, do nothing, and collect profit.
In the context of domains, a domain squatter would be more like someone who holds and keeps a domain without any intention of actually using it or taking it knowing someone else might have use of that exact name, they just want to sell it later for more money. In real-life I agree with your definition though, I guess it not 100% comparable, bad name that got stuck.
There are domain name investment companies, yes. HugeDomains owns over 5 million domains!
As for product companies, like Ramp, buying domains as an investment is generally a terrible idea. Domains are not liquid, the (hypothetical) $5 million that Ramp paid is not a price Ramp could achieve when selling the domain tomorrow. The previous owners owned Router.com for over 20 years before Ramp came along, willing to pay a lot of money for it.
Sometimes a company will buy a domain, use it, shut down whatever it was, then end up with the domain on their balance sheet for years, even decades, before someone tracks them down and offers to buy it. In those cases, the company can make money off the domain, but for every time that happens, there’s thousands more domains on balance sheets that nobody wants to pay a good amount for.
More broadly, domain name investors make their money by buying lots of domains and holding them over the long term, making their money from the very small number that sell. Part of why good domains are so expensive is that 1 sale needs to cover the cost of hundreds of even thousands of other domains that don’t sell.
Yes, that’s how you know they’re rich rich. They’re so rich they don’t care about having a website. The reputable brokers know who they are, the serious buyers will get a message to them.
It’s also so strange how they all play the exact same game. Everybody creates pretty much the same products. Every software company will have a chatbot, an agentic harness, a model router, and whatever will be the next trend. It’s not an ecosystem where we see lots of different ideas with some crazy, some failing, some succeeding, it really feels like it’s a single same bet and all SF companies are sort of acting like one homogeneous meta company
Same happened at the start of the cloud era; everyone created a YT competitor, Twitter clones, MySpace clones, a cloud platform to host on (though it was just a wrapper around AWS like wrappers we see around OpenAI [looking at you Heroku])
Big tech emails leaked during a trial over a decade ago; C-level types agreed they need not truly compete and see each other as one big company.
Screen based filter bubble living has kept the digital generation ignorant; it's all built on the same physical principles. No point pretending, at their level, they're doing anything truly different
"Router itself stores model inputs, outputs, and metadata and uses this data to improve the service, but users may turn this off in settings."
Why is it that every AI-product think dark UX with default data harvesting of proprietary data is a great idea? I mean it's rhetorical, stealing for training is the norm, but it's still just as disappointing every time I see it.
Indeed, probably not cheap. Ironically, they see themselves as the ones that should help others "stop overpaying for things" and "help companies make better spending decisions":
> Wait, Ramp bought router.com? We did. It was either this or spend the next year spelling out a longer URL on podcasts. More importantly, we build tools that help companies make better spending decisions and stop overpaying for things.
Indeed. Also, ramp.com is shorter than router.com, so I don't really understand the "spelling out a longer URL on podcasts". I guess router.com is just clearer for customers?
The fact that they put "every model" repeatedly, then don't actually support even close to "most mainstream models", makes me highly doubtful their service will be more honest and forthcoming once I'm signed up.
I wouldn't read too much into it, that was probably a big rush job once the Stripe / openrouter news broke.
> that was probably a big rush job once the Stripe / openrouter news broke
Yeah, this was my hunch too, which honestly doesn't bode great for the future, but I'm always up to being pleasantly surprised.
Is there an expectation that OR customers will jump ship due to something about Stripe, or about acquisitions in general?
let us know which "mainstream" models you feel like we are missing support for! https://docs.router.com/supported-models
GLM 5.3 (maybe because not open weights)
Otherwise I think they just meant chinese models not mainstream models.
Like Hy3, Mimo, etc. Also gemini models but not sure if they are mainstream anymore.
I believe you are the first to request these
Save for GLM 5.3 - coming soon once it percolates through our providers.
Muse Spark, Gemini, mistral and something from cohere would be nice
That said the current models ain't bad!
Gemini models are coming soon!
link confirms exactly the complaint of the parent poster
Judging by the responses, I don't think they understood my point at all.
I just want to know how much the domain cost.
and who owned it. Cisco, from the first Internet boom could have wanted that.
Less than $7B.
can confirm, was less than $7B
story here https://domaininvesting.com/ramp-acquired-router-com/
But no price.
The domain was previously owned by Innovation HQ (innovationhq.com) who own tens of thousands of domains, some of the best in the world (they sold robot.com not long ago, owns/runs whatismyip.com). Innovation were asking ~$5 million for it earlier this year. Given the huge money flowing into AI router projects, I expect Ramp paid very close to asking, or perhaps a little more. So, $5 million +/- $1 million is my guess. I assume Ramp will announce the price in a few weeks to generate some more press for the project.
I wonder if more companies use web domains as "investments" aswell
Domains investors (or squatters, as one may call them) have been around since the beginning of domains.
I think you mean landlords, not squatters. Squatters are the people who occupy owned property without permission, landlords are the people who hold property, do nothing, and collect profit.
In the context of domains, a domain squatter would be more like someone who holds and keeps a domain without any intention of actually using it or taking it knowing someone else might have use of that exact name, they just want to sell it later for more money. In real-life I agree with your definition though, I guess it not 100% comparable, bad name that got stuck.
Domain trolls. Similar to patent trolls that buy patents only to resell or license them.
There are domain name investment companies, yes. HugeDomains owns over 5 million domains!
As for product companies, like Ramp, buying domains as an investment is generally a terrible idea. Domains are not liquid, the (hypothetical) $5 million that Ramp paid is not a price Ramp could achieve when selling the domain tomorrow. The previous owners owned Router.com for over 20 years before Ramp came along, willing to pay a lot of money for it.
Sometimes a company will buy a domain, use it, shut down whatever it was, then end up with the domain on their balance sheet for years, even decades, before someone tracks them down and offers to buy it. In those cases, the company can make money off the domain, but for every time that happens, there’s thousands more domains on balance sheets that nobody wants to pay a good amount for.
More broadly, domain name investors make their money by buying lots of domains and holding them over the long term, making their money from the very small number that sell. Part of why good domains are so expensive is that 1 sale needs to cover the cost of hundreds of even thousands of other domains that don’t sell.
>InnovationHQ.com
Their site is a totally blank HTML page, it seems?
Yes, that’s how you know they’re rich rich. They’re so rich they don’t care about having a website. The reputable brokers know who they are, the serious buyers will get a message to them.
That's funny I thought it would be an empty response, but they explicitly hand-coded it like that.
Also -- offtopic but I think it's funny -- for conformance, you need a doctype and a (non-empty) title. So the above isn't valid, but this is:
I guess this is the world we live in now. All companies no matter which sector they belong to must buy AI crap.
A few years ago, companies buying random companies from a completely different sector was frowned upon by stakeholders
It’s also so strange how they all play the exact same game. Everybody creates pretty much the same products. Every software company will have a chatbot, an agentic harness, a model router, and whatever will be the next trend. It’s not an ecosystem where we see lots of different ideas with some crazy, some failing, some succeeding, it really feels like it’s a single same bet and all SF companies are sort of acting like one homogeneous meta company
Same happened at the start of the cloud era; everyone created a YT competitor, Twitter clones, MySpace clones, a cloud platform to host on (though it was just a wrapper around AWS like wrappers we see around OpenAI [looking at you Heroku])
Big tech emails leaked during a trial over a decade ago; C-level types agreed they need not truly compete and see each other as one big company.
Screen based filter bubble living has kept the digital generation ignorant; it's all built on the same physical principles. No point pretending, at their level, they're doing anything truly different
Mirror neurons are a helluva drug: https://en.wikipedia.org/wiki/Mirror_neuron
Society evolves slowly as generational churn is not fast
Interesting parallel
we (Ramp) didn't buy anything, we built a model router in house!
"create a cloudflare worker that proxies requests to any number of neoclouds. make no mistakes."
Ramp showed up late into a market and have had more success than almost every company who tried before them. They seem to have some execution skill.
Why is a bank selling llm inference?
To juice their own valuation as an "AI Company"
Routing requests to the ideal model? that's some big talk right there.
check out some of our work in this area
https://engineering.ramp.com/post/thompson-sampling-model-ro...
https://x.com/RampLabs/status/2087163448513765449
Why do these Router companies never have BAAs? They can easily get them with the vendors, why not sign with the customer?
because they would be meaningless words on trust me bro paper.
It can't be a coincidence that this dropped today, the same day news dropped about the Stripe acquisition of OpenRouter
Same day? The acquisition has been public knowledge for at least a week.
I just bought rowtr.ai, it should be ready by the time this one is acquired
Do you see Router eventually taking customer entitlements or credit balances into account, or is the plan to keep it focused on model selection?
Incredible timing.
Once again a common word for an established technology, (network routers), has been hijacked by a techbro company. How's your NFT collection going.
they really buried the lede on which models are supported, then the link is broken as of now to the list (https://app.router.com/supported-models)
thanks for the heads up! fixed!
"Router is not available in your region yet"
Hmmm
"Router itself stores model inputs, outputs, and metadata and uses this data to improve the service, but users may turn this off in settings."
Why is it that every AI-product think dark UX with default data harvesting of proprietary data is a great idea? I mean it's rhetorical, stealing for training is the norm, but it's still just as disappointing every time I see it.
Paypal, where's your AI model abstraction layer?
router.com! Cool domain though that must have been expensive :P
> that must have been expensive
Indeed, probably not cheap. Ironically, they see themselves as the ones that should help others "stop overpaying for things" and "help companies make better spending decisions":
> Wait, Ramp bought router.com? We did. It was either this or spend the next year spelling out a longer URL on podcasts. More importantly, we build tools that help companies make better spending decisions and stop overpaying for things.
Indeed. Also, ramp.com is shorter than router.com, so I don't really understand the "spelling out a longer URL on podcasts". I guess router.com is just clearer for customers?
It's certainly cheaper than buying OpenRouter for ~7B
Money can be exchanged for goods and services. Optimizing your saas/vendor/flights/hotel spend frees up money for domains!
Looking forward to their inclusion on models.dev.
That's some domain;
Blog post: https://ramp.com/blog/router-launch
Thats basically a direct response of OpenRouter by Stripe? ahaha
US only, right?
for now, yes