Forget the model wars. The router wars just began.
Model triage is becoming one of the most important skills for the AI-native developer. I’ve argued all summer that the people getting the most out of frontier models are the ones disciplined enough not to run the best model by default. On Wednesday, Stripe and Ramp validated that idea 70 minutes apart: Stripe bought OpenRouter, and Ramp released its internal router.
Bloomberg puts the OpenRouter price tag above $7 billion, and Axios says it’s more than $8 billion in cash and stock. Stripe has not released the terms, so the details remain fuzzy.
While the acquisition made headlines, the architecture is the story. For the last couple of years, picking a model was something written into an application, a string in a config file, and swapping models took some work. A router changes that workflow. Stripe bought the layer, and Ramp built it. Both are betting their existing relationships give them a unique wedge to own this critical layer in the new AI stack.
WeAreDevelopers welcome reception with The New Stack and Dynatrace
Elevated bites. Big ideas. One incredible venue. Join us in San Jose on September 23, 6-8 PM for an evening exploring one of the Bay Area’s best tech museums.
Mingle with the brightest minds and VIPs in tech
Meet The New Stack’s editorial team in person
Explore The Tech Interactive museum after hours
Space is limited, so get your name on the list soon!
Doist CTO Gonçalo Silva doesn’t pretend to know where AI is headed. The pace of change has made it harder to commit to one roadmap, model provider, or way of working. He is betting that generative models should translate intent into predictable code rather than remain in the execution loop. In this episode, he joins The New Stack's Frederic Lardinois to discuss how the company behind Todoist is approaching AI without simply adding a chatbot or shipping every possible AI feature.
Smarter alerting at scale: Live OpenSearch demo on PPL & unified alerting
Most observability platforms force a tradeoff: alert on everything and pay exponentially, or scale back and accept the blind spots. This upcoming live session shows how two OpenSearch capabilities close that gap: no licensing tiers, no ingestion ceiling. Spots are limited for this webinar; register today to claim yours!
What breaks in your retrieval layer when hundreds of agents hit it at once
Your retrieval layer works great until hundreds of agents hit it at once. Explore the failure modes that show up when retrieval has to serve concurrent agent workloads instead of human ones: latency stacking, stale context, and relevance drift under load.
Human review vs. verified pipelines: What catches bugs in the age of AI code
Code review wasn't built for this volume. AI agents now write faster than any team can check, pull requests up nearly 2x, bugs up 54%. Join us live to see what actually catches bugs when review can't keep up.
This event brings together the developers, researchers, and enterprises building the next generation of AI agents, from core architecture to the protocols that make agent systems interoperable. Through technical sessions and real-world case studies, the conference explores how teams are building reliable, scalable agent systems in production. The New Stack will be there covering all the action.
.png?height=72&upscale=true&name=The%20New%20Stack%20logo%20-%20Reversed%20(2).png){kind=small}
