Everyone has been asking me why Ramp is building a router
We started by helping companies manage spend.
As more of their money goes to buying intelligence, our job is to help 85k customers decide between models, too
Excited about the progress so far, but still a lot of work ahead
We started by helping companies manage spend.
As more of their money goes to buying intelligence, our job is to help 85k customers decide between models, too
Excited about the progress so far, but still a lot of work ahead
How do model routers compare to individual models on real agentic tasks? Introducing the Vals Router Index.
Routers promise to balance quality and cost across a fast-moving model frontier. We put that promise to the test with an independent evaluation of five routers (OpenRouter Router, Jev, Fugu Max) against every single model on the Vals Index, using the same agents, ~1,250 tasks and graders across finance, coding, law and tax.
We found that routing didn’t consistently deliver better value. Every router was beaten on both score and cost by at least one single model.
1
4
66
Ramp Router is literally free money if you're spending 5.5-8% on top of list prices with OpenRouter
I'll personally come to your offices in NYC or SF to get you onboarded
I'll personally come to your offices in NYC or SF to get you onboarded
Rolling something out to an entire engineering org means dealing with the less fun stuff: budgets, billing, data controls, and more.
We just shipped a bunch to make it easier for Admins to bring teams onboard:
1. Org management: Invite engineers and assign co-admins. Everyone a one-line install for their coding agents, and routing strategies for what they're building.
2. Policies: Configure which models and providers people can use, and set data retention policy at the org and individual levels.
3. Spend controls: Set budget caps for the org, each individual, and each API key. See spending across the org.
4. Key management: See everyone’s key, share provider keys across the org, let personal keys take precedence, and control fallback to Ramp's shared key.
5. Payments: Prepay with a card or pay monthly based on usage.
6. One Ramp account, one bill: Already use Ramp? Sign in with your existing Ramp account and put Router usage on your monthly Ramp statement.
We want every engineer to have the best setup for their work, without making finance and IT nervous about what it's all going to cost.
10
6
166
I’m hiring FDEs to work on Ramp Router at AI pilled companies
Current/former founders who can build and sell will be unstoppable here
Ideally you are:
- Obsessed with something to an unreasonable degree
- Brimming with crazy ideas
Someone on our growth team put a Ramp ad . Last night we ran a one-night-only Broadway musical about paying bills
- High agency and actually follow through to execution
Email or DM me if interested
Current/former founders who can build and sell will be unstoppable here
Ideally you are:
- Obsessed with something to an unreasonable degree
- Brimming with crazy ideas
Someone on our growth team put a Ramp ad . Last night we ran a one-night-only Broadway musical about paying bills
- High agency and actually follow through to execution
Email or DM me if interested
26
5
290
Ashwin benchmarked Jev as a replacement for LLM-based reranking in Ramp's accounting product, and the results are very promising
Jev matches current accuracy on GPT-5.6 Luna, but decreases tail latency by 10x to 300ms at 3x lower cost
We're excited to get this into production on as we stop running into rate limits :)
Jev matches current accuracy on GPT-5.6 Luna, but decreases tail latency by 10x to 300ms at 3x lower cost
We're excited to get this into production on as we stop running into rate limits :)
27
15
193
Neel moved Ramp’s structured receipt extraction from Gemini 3.5 Flash-Lite to DeepSeek-V4.1-Flash:
- 70% cheaper (~$40k/month saved)
- about the same accuracy
- about the same latency
Running through router.com
- 70% cheaper (~$40k/month saved)
- about the same accuracy
- about the same latency
Running through router.com
23
9
429
Monitor and control your AI spend on every provider on router.com. Our early users save 40% on average.
Every week, the price-intelligence-latency frontier shifts, and we expect this trend to continue. Tradeoffs between latency, reasoning, cost, service tier, open e models are shifting constantly.
Router sends every request to the model that's actually best for the task and helps you control what tokens you buy. We benchmark it against real work: ~40% lower cost for the same outputs.
Today we're opening it to everyone.
Two lines of code or just change your base URL. No @tryramp account needed.
Free through 2026, first $26 on us.
Get an API key today at router.com
Every week, the price-intelligence-latency frontier shifts, and we expect this trend to continue. Tradeoffs between latency, reasoning, cost, service tier, open e models are shifting constantly.
Router sends every request to the model that's actually best for the task and helps you control what tokens you buy. We benchmark it against real work: ~40% lower cost for the same outputs.
Today we're opening it to everyone.
Two lines of code or just change your base URL. No @tryramp account needed.
Free through 2026, first $26 on us.
Get an API key today at router.com
148
107
1,376
Everyone is underestimating NVIDIA's ability to innovate up and down the AI stack
We implemented NVIDIA Switchyard on our internal SWE-Bench and cut costs by 58%
It's live in Ramp Router today - we're actively removing people from the waitlist while we manage load!
We implemented NVIDIA Switchyard on our internal SWE-Bench and cut costs by 58%
It's live in Ramp Router today - we're actively removing people from the waitlist while we manage load!
We tested NVIDIA NeMo Switchyard’s stage router for coding agents in Ramp SWE-Bench.
Routed agents showed comparable performance to single-model controls while substantially reducing costs and runtime.
5
7
117
there’s no substitute for being obsessed and making sure every token counts 🫡
There is no substitute for looking at the data. Manually inspecting failure modes is how you discover the shape of the problem.
Great AI products are built by teams who set up their infra + instrumentation to make looking at data easy.
@austospumanto @bcherny @rahulgs
1
2
108
Today we’re launching Ramp Router.
3 years ago, we built an internal LLM router at @tryramp that powers AI products for 70,000 customers.
Back then it was mostly about saving money.
Now it feels obvious: the best model changes constantly. GPT, Claude, Gemini, Grok, Qwen, rices and capabilities move every week.
So we’re opening up access to everyone.
One OpenAI-compatible endpoint. The right model for every request. Lower cost without rewriting your app.
Reserve access to use it.
3 years ago, we built an internal LLM router at @tryramp that powers AI products for 70,000 customers.
Back then it was mostly about saving money.
Now it feels obvious: the best model changes constantly. GPT, Claude, Gemini, Grok, Qwen, rices and capabilities move every week.
So we’re opening up access to everyone.
One OpenAI-compatible endpoint. The right model for every request. Lower cost without rewriting your app.
Reserve access to use it.
215
201
1,819
just found out there's a Pi extension that's using @tryramp's SWE-Bench as a model router
every harness will need something like this
every harness will need something like this
Which companies are building in the open about their benchmarks?
- Ramp
- Databricks
- DoorDash
1
1
60
My dad came to America in 1986 at 18 years old and worked nights/weekends to pay his way through college.
He was a cashier at a local convenience store, then worked in a printing factory, then the NYCDOT, then took the leap to start his own printing company out of our garage ey’ve been operating for the last 25 years and if you’ve ever been to Whole Foods, Target, or Walmart you’ve seen his labels on their products.
He’s one of the most patriotic people I know and loves America. I feel incredibly grateful for the opportunities this country has given our family. Happy 250th 🇺🇸
He was a cashier at a local convenience store, then worked in a printing factory, then the NYCDOT, then took the leap to start his own printing company out of our garage ey’ve been operating for the last 25 years and if you’ve ever been to Whole Foods, Target, or Walmart you’ve seen his labels on their products.
He’s one of the most patriotic people I know and loves America. I feel incredibly grateful for the opportunities this country has given our family. Happy 250th 🇺🇸
18
14
340
the best product we built at ramp was the culture - this is the best articulation i've seen yet from krish @krrishd
Today, Ramp raised $750M at a $44B valuation.
Last time we grew this fast, we were 1/20th the size.
For 2000 years, business was built on two pillars. Today, a third: intelligence.
It’s your least governed cost. It’s also your single greatest opportunity.
13
27
628
Uber's CTO told @LauraBratton5 that AI coding tools—particularly Anthropic’s Claude Code—has already maxed out its 2026 AI budget 📈
“I'm back to the drawing board, because the budget I thought I would need is blown away already,” Neppalli Naga said.
theinformation.com/newsletters/ap…
1
2
72
NYC is so back
CALLING ALL FOUNDERS.
You're invited to an evening of conversation and community with @FoundersPodcast, @patrick_oshag, and @eglyman.
4/7. San Francisco. RSVP: partiful.com/e/OR5om90Lm82s…
2
0
49
Excited to finally give my assistant Aria access to accept bribes via Ramp CLI 🤑
Today, we're releasing Ramp CLI to let agents manage your company's finances.
50+ tools across cards, bills, expenses, travel, and approvals. Fewer tokens than MCP, and comes with pre-built skills like receipt compliance and agentic purchasing.
0
0
25
on AI budgets and token cost:
- no one knows how to estimate/project this spend for the quarter, let alone the year
- finance teams are taking screenshots of admin usage dashboards
- AI token cost will outpace salaries
- OAI/Ant/Google are not incentivized to help you
- it's least understood line item in every company's budget right now
- no one knows how to estimate/project this spend for the quarter, let alone the year
- finance teams are taking screenshots of admin usage dashboards
- AI token cost will outpace salaries
- OAI/Ant/Google are not incentivized to help you
- it's least understood line item in every company's budget right now
18
5
123
Did this and now every turn feels like I'm gambling
@clockworkwhale I disagree. Hide the token stuff and just make it transparent the price of every turn. It makes sense to me
3
0
43
A lot of talk about AI diffusion and how long things will take, but it has been unreal to experience it up close with the Ramp team every day
@rahulgs @EliBBlock
@rahulgs @EliBBlock
Introducing Claude Code Security, now in limited research preview.
It scans codebases for vulnerabilities and suggests targeted software patches for human review, allowing teams to find and fix issues that traditional tools often miss.
Learn more: anthropic.com/news/claude-co…
3
2
47
despite all my persistent selling on @tryramp only seeing the rizzler in the flesh could convince my friends to start an application asap
many such cases today
many such cases today
2
2
89
Software Is Changing (Again) and we’re excited to push the bleeding edge of it at Ramp.
Shout out to @cj_enr, Will Koh, @rakesh_nori, @fel9z, and the whole team for crushing this launch.
Full blog post on all of our learnings below.
Shout out to @cj_enr, Will Koh, @rakesh_nori, @fel9z, and the whole team for crushing this launch.
Full blog post on all of our learnings below.
Finance teams waste hours chasing receipts and answering weird policy questions. Now they don’t have to.
Introducing our first @tryramp AI agents. Upload your company policy and put the bullshit work on autopilot so your finance team can actually do finance.
3
2
63
Cursor + Claude 3.5 Sonnet is really good at low complexity API “plumbing” that no one likes doing
Tried this exact setup with the Splitwise API a couple weeks ago and it worked flawlessly (with accurate currency conversion!)
Tried this exact setup with the Splitwise API a couple weeks ago and it worked flawlessly (with accurate currency conversion!)
New gold standard for developer docs:
If you can @ the url in cursor and get a finished integration in your codebase
1
0
24
As the intern that took these pitch meeting notes 5 years ago (1:45), it was funny to look back on them and see how much of it has come true:
-Gmail/outlook integration
-Vendor management
-Price intelligence
@tryramp is a special place because of this simple, relentless focus.
-Gmail/outlook integration
-Vendor management
-Price intelligence
@tryramp is a special place because of this simple, relentless focus.
After wrapping our 20th board meeting last week, @rabois and I joined @HarryStebbings to talk through our partnership in building @tryramp. Harry did a great job and I think this turned into one of the more info dense, useful podcasts for founders we've made. Hope it's useful!
1
0
85
My first time seeing a Ramp ad in person this morning on my walk to the office. New round + our marketing team’s timing is locked in 🔐
Ramp is completing a $150 million Series D-2 funding round, raising the company’s valuation to $7.65 billion.
This round was co-led by @khoslaventures and @foundersfund, with additional new investors @sequoia, @GreylockVC, and @8vc. Our existing investors @ThriveCapital, Capital, D1 Capital Partners, @Lux_Capital, @ICONIQGrowth, Definition Capital, @contrary, and more are doubling down. Thank you to our partners and supporters for believing in our vision.
This isn’t just a financial milestone. It's a recognition of every customer who has trusted Ramp to redefine their financial operations – and a bet on every customer who will trust us in the future.
More details below…
0
0
63
Throwback T̶h̶u̶r̶s̶d̶a̶y̶ Monday
One quote from @genejaelee that I still repeat to this day: "In the long run, customers will care more about how you handled and fixed their bugs ASAP more than the fact that they encountered it in the first place."
Thank you to our customers!!
One quote from @genejaelee that I still repeat to this day: "In the long run, customers will care more about how you handled and fixed their bugs ASAP more than the fact that they encountered it in the first place."
Thank you to our customers!!
5 years and over 25,000 businesses running on Ramp. 🚀
Today we’re celebrating the incredible companies of all shapes and sizes who do more on Ramp.
Together, we’ve saved them over $1 billion and over $10 million hours.
Here are just a few of their stories 🧵
2
1
57
Thank you to the team at TreeHacks for inviting me back to campus this past weekend to speak about my journey from hackathons to @tryramp in the same classroom where I took my last CS final!
2
8
109
Can confirm the 2021 version of this will be even more 🤯
Help us build it from idea to spec to ship: ramp.com/careers
DMs open
Help us build it from idea to spec to ship: ramp.com/careers
DMs open
This is a fun one.
Ramp just became the fastest unicorn in NYC history... twice, within a day.
@tryramp's @eglyman & @karimatiyeh let me behind the scenes to share what it's like to raise at $1.1bn from D1 & a $1.6bn from Stripe, back-to-back.
notboring.co/p/ramps-double…
0
1
24
Reading iOS 14 API documentation and if you squint, you can easily see the emerging building blocks for Apple's future AR glasses: spatial audio for AirPods, App Clips, NearbyInteraction, ARKit 4 🤯
0
3
31
4 years ago, during my freshman year at Stanford, I saw @eglyman and @karimatiyeh on @justinkan's Snapchat story and sent a cold email asking if Paribus was looking to hire interns for the summer…
7
20
261
@nikillinit Eventbrite and OpenTable started with stand-alone supply side software and worked on the demand side later news.greylock.com/paving-the-way…
0
0
2
Hey @tacobell can some new Stanford grads cop a sponsorship? #Stanford19 @aneeshpappu @RohunSaxena @abhikulgod
2
3
50