# RAW — Servers for AI
> 100% API. Create servers. Scale. 100× cheaper, 100× faster, 100× better than AWS.

Written for AI agents. Humans: https://rawhq.io/docs

One server = one job. Do not cram Postgres + Redis + GPU + API onto one VM. Spin up more.

- Full playbook: https://api.rawhq.io/llms-full.txt
- Live JSON: GET https://api.rawhq.io/help
- Also: https://rawhq.io/llms-full.txt
- Human docs: https://rawhq.io/docs
- Pricing: https://rawhq.io/pricing
- CLI: https://www.npmjs.com/package/rawhq

## Do this (in order)

1. Create account: `POST https://api.rawhq.io/signup` `{"email":"..."}` → save `api_token`.
2. If deploy returns `"setup": true` → `POST /billing/setup` (auth) → user opens `url` (card, $0). Retry deploy.
3. Deploy: `POST /deploy` with Bearer token, always pass `type` + `region`. Charged **before** the VM starts.
4. Deploy **more** servers: Postgres, Redis, GPU, workers — each its own `POST /deploy`.

## Default commands

```
npx rawhq init
npx rawhq deploy -t raw-5 -n api -r eu
npx rawhq deploy -t raw-5 -n postgres -r eu
npx rawhq deploy -t raw-5 -n redis -r eu
npx rawhq deploy -t raw-gpu-44 -n llm -r eu
```

```
POST https://api.rawhq.io/deploy
Authorization: Bearer $RAW_API_TOKEN
Content-Type: application/json

{"type":"raw-5","region":"eu","name":"api","apps":["docker","nodejs"]}
```

## Hard rules

- Base URL: `https://api.rawhq.io`
- Auth: `Authorization: Bearer <api_token>`
- CPU `raw-5` / `raw-4x` / `raw-8x` / `raw-16x`, ARM, and **all GPUs**: region `eu` or `eu-fi` only.
- Dedicated RAM (`raw-2d`…): `eu`, `eu-fi`, `us`, `us-west`, `sg` (Singapore costs more).
- Never reuse one payment to deploy a bigger SKU. Each server is a new charge.
- If `checkout: true`, `POST /checkout` `{ "type", "region" }` then retry deploy.
- GPU: monthly + one-time setup, billed before hardware is reserved. Ready in 1–48h.

## Prices (GET /types is source of truth)

CPU EU: raw-5 $9 · raw-4x $13 · raw-8x $22 · raw-16x $40
ARM EU: raw-4 $15 · raw-8 $29 · raw-16 $55
Dedicated: raw-2d $58 · raw-4d $116 · raw-8d $186 · raw-16d $368 · raw-32d $708 · raw-48d $1130
GPU EU: raw-gpu-44 $304/mo + $150 setup · raw-gpu-131 $1555 + $778 · raw-gpu-131x $2334 + $1168 · raw-gpu-131p $3001 + $1501

## Live catalog

GET https://api.rawhq.io/types
GET https://api.rawhq.io/regions
GET https://api.rawhq.io/apps

## Product pages

- Home: https://rawhq.io
- Product: https://rawhq.io/product
- About: https://rawhq.io/about
- Security: https://rawhq.io/security

## Compare

- RAW vs AWS: https://rawhq.io/blog/aws-vs-bare-metal
- RAW vs DigitalOcean: https://rawhq.io/blog/digitalocean-alternative
- GPU cloud: https://rawhq.io/blog/gpu-cloud-comparison
- GPU servers for AI: https://rawhq.io/blog/gpu-server-guide
- Self-host LLM (vLLM / Ollama): https://rawhq.io/blog/self-host-llm
- Fly.io alternative: https://rawhq.io/blog/fly-io-alternative
- Vercel alternative: https://rawhq.io/blog/vercel-alternative
- Railway alternative: https://rawhq.io/blog/railway-alternative
- Render alternative: https://rawhq.io/blog/render-alternative

## Contact

RAW <hey@rawhq.io>