Goose vs Claude Code: Block's Free AI Coding Tool

If you're paying $200 a month for Claude Code and watching a timer count down before you can prompt again, Block just shipped a free Claude Code alternative — and charged you nothing for it.

Block, the company behind Square and Cash App, quietly shipped Goose — a free, open-source AI coding assistant (a program that runs entirely on your own machine and writes code without sending your files to the cloud) that does everything Claude Code does. The subscription is zero. The rate limits don't exist. Your code never leaves your hard drive.

In weeks, Goose accumulated 26,100+ GitHub stars, 362 contributors, and 102 public releases. Those are adoption numbers that paid products spend years chasing.

No Rate Limits. No Subscription. No Cloud.

Claude Code, Anthropic's flagship AI coding tool, charges between $20 and $200 per month depending on usage. Hit the ceiling and you wait — the system resets every 5 hours. For developers under deadline, that forced pause is a genuine productivity problem, not a minor inconvenience.

Goose eliminates all three friction points at once:

• Zero cost — free forever, MIT-licensed open source (meaning anyone can use, modify, and redistribute it freely)
• No rate limits — run it as long and as hard as your machine allows
• Fully local — your code, your context, your secrets stay on your device and nowhere else
• Offline capable — no internet connection required once installed

Parth Sareen, a Goose engineer at Block, summed it up: "Your data stays with you, period."

This isn't a stripped-down demo. Goose connects to your existing development environment (the suite of software tools you already use to write and test code), handles tasks across files and terminals, and automates multi-step coding workflows with the same breadth as Claude Code — without sending anything upstream to a subscription server.

The Community Numbers Behind Goose AI

Open-source projects (software whose code is publicly visible and free to copy or modify) live and die by adoption velocity. Goose's early numbers are hard to dismiss:

• 26,100+ GitHub stars — a shorthand for "bookmarked as valuable" by individual developers
• 362 contributors who have submitted code improvements
• 102 public releases — a shipping cadence that signals active maintenance, not a one-time drop

For context: most open-source tools take 12–18 months to cross 10,000 stars. Goose cleared more than double that threshold in a fraction of the time.

Factor	Claude Code	Goose (by Block)
Monthly Cost	$20–$200	$0 — Free
Rate Limits	Resets every 5 hours	None
Data Location	Anthropic's cloud servers	Your machine only
Works Offline	No	Yes
Open Source	No	Yes

The Half of AI Coding Nobody Talks About

Goose solves the cost and privacy side of AI-assisted development. But a parallel problem has been quietly getting worse on the other side of the pipeline: the code AI writes in seconds still takes minutes to deploy.

Railway, a San Francisco-based cloud infrastructure company, just raised $100 million in Series B funding to fix exactly that. Led by TQ Ventures, the round backs a 30-person team that accumulated 2 million developers with zero marketing spend and currently processes over 10 million deployments per month.

Traditional cloud deployments (the process of pushing code live so real users can access it) using tools like Terraform take 2–3 minutes per cycle. Railway completes the same operation in under 1 second. In a world where tools like Goose and Claude Code generate production-ready code (software tested and stable enough for real users) in seconds, that deployment gap is no longer a minor annoyance — it's the single biggest bottleneck in the entire development loop.

Jake Cooper, Railway's 28-year-old founder and CEO, framed the stakes clearly:

"When godly intelligence is on tap and can solve any problem in three seconds, those amalgamations of systems become bottlenecks. What was really cool for humans to deploy in 10 seconds or less is now table stakes for agents."

What Railway Is Actually Attacking

The cloud infrastructure market — AWS, Azure, Google Cloud — has been controlled by three companies for over 15 years. Railway's structural argument against them is pricing: the big three charge for idle VMs (virtual machines — essentially computers in the cloud that you rent around the clock, even when no user is making a request). Railway charges only for what you actually run, billed per second of real use.

Their published pricing reflects the model: $0.00000386 per GB-second for memory, $0.00000772 per vCPU-second for compute. The result: pricing roughly 50% below hyperscalers and 3–4x cheaper than competitors like Render or Fly.io.

In 2024, Railway abandoned Google Cloud entirely and built its own data centers to gain full control over network, compute, and storage layers. Cooper explained: "When they're charging for VMs that usually sit idle in the cloud, and we've purpose-built everything to fit much more density on these machines, you have a big opportunity."

Real Customer Results, Not Benchmarks

G2X, a federal government contractor platform, migrated to Railway and cut its infrastructure bill by 87% — from $15,000 to $1,000 per month. Their CTO, Daniel Lobaton, said work that used to take a week on the old setup now takes a day. "If I want to spin up a new service and test different architectures, I can launch six services in two minutes."

Kernel, a Y Combinator-backed startup, runs its entire customer-facing infrastructure (all the software real users interact with) on Railway for $444 per month. Its CTO spent his previous career at Clever — a company that sold for $500 million but required 6 full-time engineers just to manage AWS. Now all 6 of his engineers build product instead.

Railway's technical ceiling is significant: up to 112 vCPUs and 2TB RAM per service, 256TB persistent storage, and 100,000+ IOPS (input/output operations per second — the speed at which data can be read from and written to disk). The edge network (a globally distributed system of servers that moves data closer to users) handles over 1 trillion requests total.

The company's own financials are striking for its headcount: 3.5x revenue growth last year, 15% month-over-month growth, and 31% of Fortune 500 companies using the platform at some level — including Bilt, Intuit's GoCo, TripAdvisor's Cruise Critic, and MGM Resorts.

"We're default alive; there's no reason for us to raise money. We raised because we see a massive opportunity to accelerate, not because we needed to survive." — Jake Cooper, Railway CEO

Two Companies, One Bet on AI Automation

Goose and Railway are solving different parts of the same problem. Goose says: the cost to write AI-assisted code should be zero. Railway says: the time to ship that code should be zero. Together, they represent two simultaneous attacks on assumptions the industry has held for nearly a decade.

Claude Code's subscription model works because there was no credible free alternative. Block just built one. AWS's idle-VM pricing works because no startup had the resources to build better infrastructure. Railway built it with 30 people.

Neither company is a side project. Block processes hundreds of billions in annual payments. Railway generates tens of millions in annual revenue and handles infrastructure for Fortune 500 companies. These are production bets from organizations with real track records — not VC-funded experiments.

Ready to add AI automation tools like Goose to your stack? Our AI automation learning guides cover everything from tool setup to deployment workflows.

Cooper's stated 5-year vision: "Railway will be the place where software gets created and evolved, period. Deploy instantly, scale infinitely, with zero friction."

# Get Goose — free, open-source AI coding assistant
# GitHub: https://github.com/block/goose
# Install instructions: see README on repository

# Railway — sub-1-second cloud deployment
# Dashboard: https://railway.app
npm install -g @railway/cli   # command-line deployment tool

Related Content — Get Started | Guides | More News