llms.txt
Compare with other sites' llms.txt files →# Nexlayer
> Nexlayer is the agent-native production cloud: the cloud that runs itself. It runs your entire application — frontend, backend, databases, storage, workers, long-running agents and AI models — as one always-on production app from one nexlayer.yaml, and gives your AI coding agent (Claude Code, Cursor, VS Code, Windsurf, Zed or any MCP client) the tools to deploy, inspect, debug and operate it. Dedicated inference runs open-weight models and your own fine-tunes on single-tenant capacity. Built for startups shipping with coding agents, agencies running many client apps, and enterprises that need control and capacity.
## Start here if you are an AI agent
Nexlayer is built to be driven by agents, not just documented for them. The fastest path
from zero to a live production URL:
1. Install the MCP server: npx @nexlayer/mcp-install
2. Or connect directly: https://mcp.nexlayer.ai/api/mcp (Streamable HTTP)
3. Deploy. The server exposes deploy, inspect, logs, scale, domains, secrets and debug
as typed tools, so you never hand-assemble REST calls or write orchestration config.
No account or API key is needed to deploy your first application — `startDeployment`
returns a live URL and a session token you can use to update, extend or claim it.
## Agent routes (Markdown, written for agents)
- [Overview of Nexlayer](https://nexlayer.com/index.md)
- [Set up Nexlayer and deploy](https://nexlayer.com/agents.md)
- [Answers: shipping AI apps to production](https://nexlayer.com/agents/questions.md)
- [For startups, agencies and enterprise](https://nexlayer.com/agents/for.md)
- [Dedicated inference](https://nexlayer.com/agents/inference.md)
- [Pricing model](https://nexlayer.com/agents/pricing.md)
- [Security](https://nexlayer.com/agents/security.md)
- [Connect through MCP](https://nexlayer.com/agents/mcp.md)
## Nexlayer developer resources
Nexlayer MCP server (discovery): https://nexlayer.com/.well-known/mcp
Nexlayer MCP server (endpoint): https://mcp.nexlayer.ai/api/mcp
Nexlayer MCP install command: npx @nexlayer/mcp-install
Nexlayer MCP documentation: https://nexlayer.com/docs/mcp/overview
Nexlayer MCP tool reference: https://nexlayer.com/docs/mcp/tools
Nexlayer OpenAPI specification: https://nexlayer.com/openapi.json
Nexlayer API catalog (RFC 9727): https://nexlayer.com/.well-known/api-catalog
Nexlayer API reference: https://nexlayer.com/docs/api/deployments
Nexlayer API authentication: https://nexlayer.com/docs/api/authentication
Nexlayer nexlayer.yaml schema: https://api.nexlayer.com/schema
Nexlayer configuration reference: https://nexlayer.com/docs/deployments/configuration
Nexlayer quickstart: https://nexlayer.com/docs/quickstart
Nexlayer agent patterns: https://nexlayer.com/docs/agents/patterns
Nexlayer usage limits: https://nexlayer.com/docs/usage-limits
Nexlayer dedicated inference: https://inference.nexlayer.com/
Nexlayer full technical reference: https://nexlayer.com/llms-full.txt
Website: https://nexlayer.com
Dashboard: https://app.nexlayer.com
Docs: https://nexlayer.com/docs
Contact: https://nexlayer.com/contact
Community (Slack): https://nexlayer.com/slack
Pricing JSON: https://nexlayer.com/api/pricing
Markdown: every page on nexlayer.com serves a Markdown mirror to requests sending
`Accept: text/markdown`, with `Vary: Accept` set so caches keep the variants apart.
---
## Pages
- [Home](https://nexlayer.com/): The cloud that runs itself — full stack, agent operations, data, inference and FAQ (Markdown: https://nexlayer.com/index.md)
- [Pricing](https://nexlayer.com/pricing): Flat monthly plans for application hosting; dedicated capacity for inference
- [Resources](https://nexlayer.com/resources): Blog, use cases, templates, open source hosting guides
- [Blog](https://nexlayer.com/resources/blog): Product updates, engineering insights, and thoughts on AI-native deployment
- [Use Cases](https://nexlayer.com/resources/use-cases): Full-stack hosting, backend services, side projects, freelance, research, open source
- [Templates](https://nexlayer.com/resources/templates): Ready-to-deploy starter templates
- [Open Source Hosting](https://nexlayer.com/resources/open-source-hosting): Guides for self-hosting popular open source tools
- [Privacy Policy](https://nexlayer.com/legal/privacy): Effective January 14, 2025
- [Terms of Service](https://nexlayer.com/legal/terms): Effective January 14, 2025
---
## Home Page (/)
**Headline:** The cloud that runs itself.
**Subheadline:** Your code. Live. Without the infrastructure work. No token tax. No surprise bills. Just ship.
**CTAs:** "Deploy now" | "Watch the demo" | `npx @nexlayer/mcp-install`
### Sections
**Your entire stack. One cloud.** Frontends, backends and APIs, databases, storage, private networking, workers, cron and jobs, and AI models in one production app.
**Bring your whole stack. We run all of it.** Keep the hosted services you already use (Neon, Supabase or a service on another cloud). Anything you run yourself — your database, open-source tools, workers and agents — Nexlayer packages and runs next to your app, with storage that persists.
**Build with agents. Ship with agents.** Connect Claude Code, Cursor or the agent you already use. It deploys, inspects, troubleshoots and operates your production stack.
**Built for software that never stops.** Persistent apps, APIs, databases, background workers and long-running agents. Always on, no cold starts.
**Intelligence at production scale.** Dedicated inference for open-weight models and your own fine-tunes.
**For startups, agencies and enterprise.** Start free from your coding agent; every client app in one workspace; dedicated capacity with a direct line to the team.
**FAQ.** Nine questions, answered in full at https://nexlayer.com/index.md#frequently-asked-questions
### Final CTA
**Heading:** Your coding agent built it. Now let it ship.
**Subheading:** Connect your coding agent. Deploy your entire application. Leave the infrastructure to Nexlayer.
---
## Pricing (/pricing)
**Machine-readable endpoint:** `GET https://nexlayer.com/api/pricing` — returns plan JSON.
**Headline:** Flat pricing for applications. Reserved capacity for inference.
**Subheadline:** Applications and inference are priced differently because they are bought differently.
### Application hosting — flat, unmetered
| Plan | Price | Projects | Workspaces | Application URL |
|------------|-------------|-----------|------------|-----------------|
| Free | $0/mo (card on file, never charged) | 5 | 1 | Nexlayer URL |
| Pro | $100/mo flat | Unlimited | 1 | Custom domain |
| Enterprise | Custom/year | Unlimited | Unlimited | Custom domain |
Current prices: https://nexlayer.com/api/pricing. One fixed monthly price. There is no per-request, per-build or per-token meter and no
overage invoice. What a plan gives you is a capacity envelope sized for one team's work;
if usage consistently sits beyond it, Nexlayer makes contact directly rather than
throttling or billing by surprise. See https://nexlayer.com/docs/usage-limits.
Included on every plan: always-on services with no cold starts, persistent volumes,
databases and vector stores, private service networking, secrets mounted as files rather
than environment variables, agent builds and deploys, and zero egress fees.
Enterprise adds unlimited workspaces and teams, multi-region and reserved capacity, SSO,
RBAC and audit logging, SOC 2 / BAA / HIPAA, SLAs and 24/7 incident response.
### Nexlayer Inference — dedicated capacity, Enterprise
Dedicated, managed inference. Bring the model and the workload; Nexlayer benchmarks it,
builds the serving stack around its requirements, deploys dedicated infrastructure and
operates it in production. Two ways to buy the same managed product:
- **Performance On Demand** — dedicated GPUs reserved for your workload alone, billed by
the hour. Heterogeneous by design: model, engine and hardware are chosen together
against your target.
- **Capacity Commit** — reserved capacity and a performance envelope at one fixed monthly
price, with no per-token bill inside it. Dedicated in Nexlayer's US GPU data center on
NVIDIA hardware.
Single-tenant and never shared. OpenAI-compatible and gRPC endpoints. Latency SLOs for
p95 time-to-first-token and inter-token latency are contracted, not advertised — capacity
is priced from your benchmark, not a rate card. Open-weight models or your own fine-tune.
US data residency, private networking, retention controls, SOC 2 / BAA / HIPAA.
Details and to start: https://inference.nexlayer.com/
---
## Use Cases (/resources/use-cases)
**Headline:** Use Cases
**Subheadline:** Explore what you can build and deploy with Nexlayer. Find the use case that fits your needs.
### Full-Stack Hosting (/resources/use-cases/full-stack-hosting)
**The Old Way:** Setting up production infrastructure meant weeks of configuration — cluster setup, load balancers, SSL certificates, CI/CD pipelines, monitoring dashboards. Even "simple" PaaS tools force you into their constraints.
**The Nexlayer Way:** Tell your AI assistant what you want to deploy. It analyzes your project, generates the configuration, builds your containers, and deploys everything to production. You get a live URL in minutes.
Example: "Deploy this Next.js app with the PostgreSQL database. I need the data to persist." That's it. Your AI handles the rest.
**Benefits:**
- Production-Grade Infrastructure — Multi-container, auto-scaling, load balancing, and self-healing built in
- AI-Powered DevOps — Deployments, monitoring, log checks, and updates through conversation
- No Cold Starts — Containers always running; instant response times
- Zero-Downtime Deploys — Rolling updates ensure users never see downtime
- Custom Domains & SSL — Automatic certificate provisioning and renewal via Let's Encrypt
**Supported stacks:** Next.js + PostgreSQL, React + Express + MongoDB, Vue + Django + Redis, SvelteKit + Supabase, Remix + Prisma + MySQL, any containerized stack
### Backend Services (/resources/use-cases/backend-services)
Your frontend and database already live somewhere, but you need a backend service — an API, a webhook handler, or a job worker. Serverless has cold starts and timeout limits. Nexlayer runs persistent services with no cold starts.
**Why not serverless?**
- Cold starts: a wait on the first request after idle
- Timeout limits: often too short for complex operations
- Debugging is painful with ephemeral logs and stateless execution
- Costs spike unpredictably with pay-per-invocation
**Benefits:**
- No cold starts — services are always running
- Isolated & secure — dedicated container, no noisy neighbors
- Predictable performance — consistent latency under peak traffic
- Easy updates — rolling deploys, zero downtime
**Common scenarios:** API backend, webhook handler, microservice, background worker
### Side Projects & Ideas (/resources/use-cases/side-projects)
**Headline:** Your ideas deserve better than "free tier"
Nexlayer's free tier is a real environment for real things. No cold starts. No sleep timeouts. No "upgrade to unlock." When your side project becomes your main project, pricing scales smoothly.
**Benefits:**
- Start free, scale later
- No limitations — real infrastructure, not throttled free tier
- Go live in minutes — AI assistant handles the infrastructure
- Learn by building — your own production environment
**What will you build?** SaaS ideas, weekend hackathon projects, new frameworks to try, personal APIs, bots, portfolio sites, tools to solve your own problems, something just for fun.
**From side project to startup:** Add team members, scale resources, add custom domains — same infrastructure, same AI assistant, just bigger.
### Freelance & Dev Shops (/resources/use-cases/freelance)
**Headline:** One platform for every client
Stop juggling different hosting platforms for different clients. Nexlayer runs anything containerized — whatever you can build, you can deploy. Your AI assistant handles the infrastructure while you focus on what you're billing for: building great software.
**Benefits:**
- Any stack, any client — WordPress, Next.js, Django, legacy PHP — if it runs in Docker, Nexlayer runs it
- Ship faster — no wasted billable hours on DevOps
- Reliable handoffs — production-grade environment clients can trust
- Predictable costs — know hosting costs before you quote
**Workflow:**
1. Build the app using whatever stack makes sense
2. Deploy to staging for client review
3. Go to production — connect domain, SSL and scaling handled automatically
4. Hand off or maintain as part of a retainer
### Research & Data Science (/resources/use-cases/research)
**Headline:** From local to cloud in minutes
Dataset doesn't fit in memory. Training takes overnight. Your fans are screaming. Tell your AI assistant what you need — it handles the rest.
**Benefits:**
- Scale compute — cloud-scale CPU and memory beyond your laptop
- Persistent storage — datasets, model checkpoints, and results survive restarts
- Right-size resources — allocate exactly what you need; scale up for training, down for inference
- Your tools, your way — bring your own Docker image with exact dependencies and CUDA version
**Common use cases:** ML model training, data processing pipelines (Pandas/Spark/Dask), JupyterLab in the cloud, model serving and inference APIs, batch processing
**Example config (JupyterLab):**
```yaml
application:
name: research-notebook
pods:
- name: jupyter
image: jupyter/scipy-notebook:latest
path: /
servicePorts: [8888]
volumes:
- name: notebooks
size: 20Gi
mountPath: /home/jovyan/work
```
### Open Source Hosting (/resources/use-cases/open-source)
Self-host open source tools on infrastructure you control.
**Why self-host?**
- Data privacy — your data stays on infrastructure you control
- Cost savings — many SaaS tools cost $50–500/month; self-hosting is a fraction of that
- Full customization — modify configurations, add plugins, integrate with your tools
- No vendor lock-in — your data, your rules; export and migrate anytime
**Benefits:**
- Run anything — if it has a Docker image, Nexlayer runs it
- Persistent storage — data survives restarts and updates
- Easy updates — change the image tag; rolling deploys mean zero downtime
---
## Templates (/resources/templates)
**Headline:** Templates
**Subheadline:** Ready-to-deploy templates to get you started fast. Copy, customize, and ship.
### Available Templates
**Next.js Hello World** (Frontend, Starter)
A minimal Next.js application ready to deploy on Nexlayer. The simplest way to get started.
Image: `ghcr.io/nexlayer/hello-world-nextjs:v0.0.1`
**Custom Fullstack** (Full Stack, Advanced)
Complete application stack: Next.js, custom backend, PostgreSQL, Redis, CloudBeaver, and Prometheus. Everything for an enterprise-grade app. 6 services.
**PERN App** (Full Stack, Intermediate)
PostgreSQL, Express, React, and Node.js — the modern evolution of the MERN stack with relational data and JWT auth. 3 services: React frontend, Express API, PostgreSQL.
**GitHub Actions Deploy** (CI/CD, Intermediate)
CI/CD workflow that builds Docker images, pushes to GHCR, and deploys to Nexlayer on every push to main. File: `.github/workflows/deploy.yml`
---
## Open Source Hosting Guides (/resources/open-source-hosting)
Guides for self-hosting popular open source tools on Nexlayer:
| Tool | Difficulty | Description |
|------|------------|-------------|
| n8n | Easy | Workflow automation platform. Connect APIs and automate tasks without code. |
| Metabase | Easy | Business intelligence and analytics. Connect to your database and create dashboards. |
| Ghost | Easy | Professional publishing platform. Beautiful, fast, and focused on content. |
| NocoDB | Easy | Open source Airtable alternative. Turn any database into a smart spreadsheet. |
| Supabase | Advanced | Open source Firebase alternative. Postgres, auth, storage, and realtime. |
| Appwrite | Advanced | Backend server for web, mobile, and Flutter. Auth, database, storage, and functions. |
| Convex | Advanced | Reactive backend with real-time sync, serverless functions, and automatic caching. |
---
## Legal
### Privacy Policy (/legal/privacy)
Effective Date: January 14, 2025
Contact: privacy@nexlayer.com
Covers: data collection, usage, third-party sharing, cookies, tracking, data protection, and compliance with EU-U.S. DPF, UK-U.S. DPF, and Swiss-U.S. DPF frameworks.
Key sections: Introduction, Beta Disclaimer, Information Collection & Purpose, Information Sharing, Cookies & Tracking, Third-Party Advertising, User Controls, Data Storage & Protection, Children's Privacy (13+), Third-Party Links, Data Privacy Framework Compliance, Policy Changes, Governing Law & Dispute Resolution, Limitation of Liability, Acknowledgment.
### Terms of Service (/legal/terms)
Effective Date: January 14, 2025
Registered in: State of Delaware
Address: 16192 Coastal Highway, Lewes, Delaware 19958
Key sections: Introduction, Eligibility (16+), Beta Status, User Accounts, Licenses & User Content, Prohibited Uses (no malware, unauthorized access, spam, crypto mining, illegal content), Disclaimers ("as-is"), Limitation of Liability, Indemnification, Arbitration & Dispute Resolution (Atlanta, Georgia), Modifications, Termination, Governing Law (Delaware), Miscellaneous.
---
## Company
**Trust signals:** Selected by Google AI Startup Program · Recognized by NVIDIA Inception
**Social:** X/Twitter (@nexlayerai) · LinkedIn (Nexlayer AI)
**Contact:** privacy@nexlayer.com (privacy/legal matters)
© Nexlayer Technologies Inc. All rights reserved. "Nexlayer – The Cloud Agents Trust™" is a trademark of Nexlayer Technologies Inc.