ReadyServer.ai

Free with your ReadyServer VPS

A private LLM API, built into your VPS

Run AI agents, RAG pipelines and chatbots on open models served inside ReadyServer's private network. OpenAI-compatible, with no API key and no per-token billing.

Get a VPS Read the quickstart

Now serving Qwen3.6-27B and Qwen3.8-27B, with more models on the way.

$ curl http://api.readyserver.ai/v1/chat/completions \
    -H "Content-Type: application/json" \
    -d '{
      "model": "qwen3.6-27b",
      "messages": [
        {"role": "user", "content": "Hello from my VPS!"}
      ]
    }'

{"choices": [{"message": {"role": "assistant",
  "content": "Hello! What are you building today?"}}]}

Why it's different

An LLM API that comes with your server

No keys to manage and no metered bill. If your app runs on a ReadyServer VPS, it can call the API.

No API key

Access comes from your VPS being on our private network. There's nothing to create, rotate or leak.

No per-token billing

The API is included with your VPS. Instead of metering tokens, capacity is shared fairly between customers.

OpenAI-compatible

Use the official OpenAI SDKs, or any tool that lets you set a custom base URL.

On the private network

Requests travel from your VPS to our inference servers over ReadyServer's private network, not the public internet.

How it works

From VPS to first response in minutes

  1. Deploy a ReadyServer VPS

    Every plan on readyserver.sg includes access to the LLM API, with nothing extra to set up.

  2. Point your code at the API

    Set the private base URL in any OpenAI-compatible SDK or tool. Any placeholder works for the API key.

  3. Build

    Ship agents, chatbots, RAG search and automations without watching a token meter.

Follow the quickstart →

Models

Open models, served with vLLM

Two Qwen models today, with more to come.

Available

Qwen3.6-27B-FP8

Answers directly for fast responses. 131K-token context, tool calling and image input.

Model details →

Available

Qwen3.8-27B-FP8

Thinks before it answers. 262K-token context, tool calling and image input.

Model details →

Coming soon

More models

We plan to add more models. Once the community forum opens, you can tell us which ones you'd like to see.

Use cases

What you can build

  • AI agents

    Give agents a model to plan and call tools, running next to the services they automate.

  • RAG and search

    Answer questions over your own documents, with both the data and the model inside ReadyServer's network.

  • Coding agents

    Run coding agents and developer tools on your VPS against an OpenAI-compatible endpoint.

  • Automation

    Summarise, classify and extract data in workflows and scheduled jobs on your server.

  • Chatbots

    Power support bots and internal assistants hosted on your VPS.

Fair share

Shared fairly, so everyone gets a turn

The API runs on capacity shared by ReadyServer VPS customers. Instead of charging per token, a fair-share policy stops any single workload from crowding out everyone else.

Read the fair use policy

  • Stream responses and set max_tokens to keep things fast for everyone.
  • Back off and retry when the service is busy.
  • When it's busy, requests are queued and customers who have used fewer tokens are served first.

Build with the ReadyServer AI community

We're opening a forum for developers working with local LLMs, agents and AI tooling. Get your VPS ready today.

Get a VPS About the community