# PMS Cloud

> PMS Cloud (PMServers) is a small cloud provider. It has VMs, containers and an OpenAI-compatible AI API. The console is https://console.pmservers.org. The API is https://api.pmservers.org.

Each link below is a raw Markdown file. All pages in one file: https://docs.pmservers.org/llms-full.txt
OpenAPI spec of the AI API: https://docs.pmservers.org/openapi.json

## Docs

- [PMS Cloud Docs](https://docs.pmservers.org/index.md): Start here. What PMS Cloud has, and where to find each topic.
- [Overview](https://docs.pmservers.org/get-started/overview.md): Accounts, projects, the region, and the free tier limits.
- [Your first VM](https://docs.pmservers.org/get-started/first-vm.md): Create an Ubuntu VM in the console and open its browser console.
- [Your first AI request](https://docs.pmservers.org/get-started/first-ai-request.md): Make an AI key and send a chat request with curl or the OpenAI SDK.
- [Create a VM](https://docs.pmservers.org/vms/create.md): VM sizes, images, disk size, name rules and the SSH key that a new VM needs.
- [Connect with SSH](https://docs.pmservers.org/vms/ssh.md): Coming soon. SSH to VMs through a jump host. Until then, use the browser console.
- [Browser console](https://docs.pmservers.org/vms/console.md): Open the serial console of a VM in the browser.
- [VM states and actions](https://docs.pmservers.org/vms/states.md): What each VM status means, and how to start, stop, restart and delete a VM.
- [Containers](https://docs.pmservers.org/containers/overview.md): Coming soon. Worker containers with no inbound traffic, no disk, linux/amd64 images, on gVisor.
- [AI API overview](https://docs.pmservers.org/ai/overview.md): The OpenAI-compatible AI API at https://api.pmservers.org/ai/v1, its endpoints and auth.
- [Models and prices](https://docs.pmservers.org/ai/models.md): The AI models, their types, and the price per 1M tokens in USD.
- [AI keys](https://docs.pmservers.org/ai/keys.md): Make, limit, rotate and revoke AI keys. Key budgets, model lists and rate limits.
- [Kev (decision model)](https://docs.pmservers.org/ai/kev.md): Use the kev decision model with POST /ai/v1/systemone to get a choice and probabilities.
- [AI errors](https://docs.pmservers.org/ai/errors.md): Error codes of the AI API, in OpenAI error format, and what to do for each.
- [S3 storage](https://docs.pmservers.org/storage/s3.md): Coming soon. S3-compatible buckets. Uploads larger than 100 MB must use multipart upload.
- [Limits](https://docs.pmservers.org/billing/limits.md): Every quota and limit of PMS Cloud on one page.
- [AI billing](https://docs.pmservers.org/billing/ai-billing.md): How AI requests are billed by token, with cache prices and the monthly budget.
- [API overview](https://docs.pmservers.org/api/overview.md): The two PMS Cloud APIs, their base URLs, auth and error format.
- [AI API reference](https://docs.pmservers.org/api/ai.md): All endpoints of the AI API at https://api.pmservers.org/ai/v1.
- [Management API reference](https://docs.pmservers.org/api/management.md): All endpoints of the management API at https://api.pmservers.org/v1.
- [For agents](https://docs.pmservers.org/agents.md): How to give these docs to an AI agent, and the machine-readable files the site serves.