We've spent the better part of 2026 building the tools companies need to put AI to work: Stack, Token Spend Management, and Glass. Today we raised $1.85B at $60B valuation. There are now 85,000 businesses on Ramp saving a combined $1B+ a month. We still have so much to build.

Why I use Glass over Codex and Claude Code: - model agnostic (try the nth new model launch without changing your setup) - cloud first (vs. frankensteined local vs. cloud distinction) - built around shared company context (not one person slinging slop grenades back and forth) And for finance/IT admins: - govern who can use which models - see and manage token spend easily - shared outputs: do the work once, instead of running redundant agents across your org

Giving everyone access to the best AI models is just the start. At Ramp, an account manager used Glass to automate capturing feature requests from customer calls. We went from logging about 300 by hand each month to capturing 20,000. Glass also prioritizes those requests and emails customers when something they asked for ships. We've sent over 5,000 of those emails. We built Glass so one person’s better way of working could help the whole company. Today, we’re opening it up to others.

There should be no distinction between "checking" and "savings" accounts in 2026 I find it sad that fintechs are still playing this arbitrage game. Mercury pays 0% on checking accounts. EFFR is 3.88%. they hide this in the product (does not show 0% anywhere). to access yield, you have to click accounts, add account, create account, click savings, click create account. and then move money there. Column that powers Mercury banking doesn't even distinguish between savings and checking

Hello world! Router team here, we're very excited to share product announcements, model releases and routing strategies with you all going forward. Join us today to access models through one API, route requests, set fallbacks, and track usage and cost.

I had the chance to work with @subawse over the past couple of years. So exciting to see her and the @llmh team take on this challenge. They are some the smartest high-agency operators out there. @alextaubman and crew are on a tear !!! Congrats. Now the real work starts 😉

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 inside a fortune cookie. 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

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

Ramp is finally live in the UK! Earlier this year, I wrote about wanting to make it easier to build a company in Europe. Very happy that more UK teams can now use what we've built. Our full-page apology is in the Financial Times. Sorry we're late 🇬🇧

How do you turn an agentic commerce demo into a production-ready agent? With Ramp, you can give an agent its own identity and spending limits, plus the ability to attach receipts, memos, and everything else finance needs to close the books. Here’s what it looks like. We’re onboarding alpha customers. Reach out via the website below for access.

We built Token Spend Management to understand AI costs and Router to use the most efficient model for the task. Now, both are available to customers. Inspect comes from the same instinct: if we need a better tool, we build it. Less than a year and over 1M+ sessions later, 75% of merged PRs at Ramp come from Inspect, our homegrown background agent. Read how Inspect changed the way we build software.

The world has changed: choosing an AI model is now a full-time job. GPT-5.6 Luna, Sol, or Terra? Claude Opus 4.8 or Sonnet 5? Gemini 3.1 Pro, Kimi K3, or GLM 5.2? Medium, high, or xhigh reasoning? Best solve rate, lowest cost, or fastest response? And next week, the answer changes again. Ramp Router keeps track so you don't have to. The right model for each task is now two lines of code:

Lulu (@lulumeservey) is the best in the world at what she does. She’s the go-to comms strategist for Silicon Valley and many of the top founders in tech trust her. She’s an advocate for going direct, has built a singular career and shares everything she’s learned. 0:00 From China to America 2:08 Learning to Read the Room 6:18 Building a Movement 7:30 Learning Comms by Doing 10:50 Anduril as an Insurgent 15:07 The Power of a Manifesto 18:47 Go Direct 26:45 Conviction Over Charisma 31:30 Why Rostra Exists 33:23 How AI Leaders Lose Trust 40:40 Tell a Better Story 47:55 Do Things That Can't Be Faked 50:55 The Return to Beauty 57:11 Originality Over Templates 1:04:02 Do What Only You Can Do 1:07:14 The Jetsons Rule 1:10:41 Trust Yourself 1:12:29 Why Humans Crave Stories Includes paid partnerships.

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.

Contrarians tend to do very well at Ramp. @shevchenkoaalex, who runs @ramplabs, is one of them. High-agency people doing hard work should have real say in what work is worth chasing. Especially when everyone else says "that sounds too hard" or "that'll never work." Those are usually the projects worth trying. If you hear "impossible" and enjoy proving people wrong through action, you're exactly who we want building with us at @tryramp: