Railway has raised $100 million in a Series B round to challenge AWS with an AI-native cloud platform built for the era of coding assistants. The San Francisco company says its infrastructure deploys in under a second and undercuts the hyperscalers on price, a direct shot at Amazon, Microsoft, and Google in the market they have dominated for more than a decade.

TQ Ventures led the round, with FPV Ventures, Redpoint, and Unusual Ventures joining. The funding lands as developers increasingly lean on AI tools that generate working code in seconds, exposing how slow traditional deployment pipelines have become.

Why deploy speed is the new battleground for AWS rivals

Railway's core pitch is speed. A standard build-and-deploy cycle running Terraform takes two to three minutes, a delay that turns into a real bottleneck when an assistant can write a feature faster than the platform can ship it. Railway claims it delivers deployments in under one second, collapsing that gap.

The company has reached two million developers without spending a dollar on marketing, and it now processes more than 10 million deployments per month while handling over one trillion requests across its edge network. Those are figures that rival far larger and better-funded competitors in the broader AI infrastructure space.

Customers report sharp cost and speed gains

G2X, a platform serving 100,000 federal contractors, measured deployments seven times faster after migrating to Railway, along with an 87 percent cost reduction. Its monthly infrastructure bill dropped from roughly $15,000 to about $1,000. That kind of result is the case study Railway wants enterprises to notice.

Building its own data centers

In 2024, Railway walked away from Google Cloud and built its own data centers, a vertically integrated bet that founder and CEO Jake Cooper says enables a differentiated build-and-deploy loop. That control also kept Railway online during recent widespread cloud outages that took down services relying on the major providers.

Owning the hardware underpins aggressive, usage-based pricing. Railway charges by the second for actual compute and does not bill for idle virtual machines:

  • $0.00000386 per gigabyte-second of memory
  • $0.00000772 per vCPU-second
  • $0.00000006 per gigabyte-second of storage

Cooper says this undercuts the hyperscalers by roughly 50 percent and newer cloud startups by three to four times. For workloads that scale up and down constantly, per-second metered billing can meaningfully lower the bill versus reserved-capacity models.

A lean team with surprisingly broad reach

Railway has reached this scale with about 30 employees and tens of millions in annual revenue. Revenue grew 3.5 times last year and continues to expand 15 percent month over month. The company says 31 percent of Fortune 500 companies now use the platform, with customers including Bilt, Cruise Critic, and MGM Resorts.

The platform supports PostgreSQL, MySQL, MongoDB, and Redis, offers up to 256 terabytes of persistent storage, and deploys across four global regions spanning the United States, Europe, and Southeast Asia. For enterprises, it provides SOC 2 Type 2 compliance, HIPAA readiness, single sign-on, and audit logs, the checklist items that gate adoption inside regulated industries.

Why it matters and what comes next

Cooper described the raise as strategic rather than necessary, calling the company profitable and choosing to accelerate. Railway plans to expand its data center footprint, grow beyond 30 employees, and build a formal go-to-market operation for the first time in its five-year history.

The bet reflects a wider shift: as AI coding tools compress how long it takes to write software, the platforms that ship that software fastest gain an edge. Whether grassroots developer enthusiasm converts into durable enterprise contracts is the open question for any challenger in cloud, a market where switching costs run deep. You can follow how these shifts play out across our ongoing technology business coverage.

Frequently asked questions

How is Railway cheaper than AWS?

Railway bills per second for actual compute used and does not charge for idle virtual machines, while owning its own data centers instead of renting from a hyperscaler. The company says this makes it about 50 percent cheaper than AWS, Microsoft, and Google.

How fast does Railway deploy compared to Terraform?

Railway says it completes deployments in under one second, versus the two to three minutes a typical build-and-deploy cycle takes with Terraform.

Who uses Railway?

Railway reports two million developers and says 31 percent of Fortune 500 companies use the platform, including Bilt, Cruise Critic, and MGM Resorts.