Docs/Get started

Osarai documentation

Rent a GPU of your own, run an open model on it, and call it with the OpenAI or Anthropic API from your code, your terminal agent or your editor.

Osarai rents a dedicated GPU for you, starts an open model on it (Qwen3.8, gpt-oss and others) and gives you one HTTPS endpoint that speaks three API formats:

• OpenAI Chat Completions at https://osarai.xyz/v1/chat/completions, the format most tools and SDKs use.
• OpenAI Responses at https://osarai.xyz/v1/responses, used by Codex CLI.
• Anthropic Messages at https://osarai.xyz/v1/messages, used by Claude Code and the Anthropic SDKs.

You pay for GPU time by the hour from a prepaid balance, not per token, and nobody else shares your machine.

#How it works

your code / agent ──HTTPS──▶ Osarai API (https://osarai.xyz)
                              │  checks your API key and balance
                              │  picks your running instance
                              ▼
                       your GPU instance (vLLM + the model)

1. Deploy a model from the dashboard. Osarai rents a GPU, downloads the model and starts it. That takes 10 to 20 minutes the first time.
2. Create an API key on the Keys page. Keys start with qck_.
3. Point your tool at the base URL below and use the key. Requests reach only your own instances.

#Base URL

Anthropic clients add /v1/messages themselves, so they take the site address without /v1.

#A first request

curl "https://osarai.xyz/v1/chat/completions" \
  -H "Authorization: Bearer $OSARAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3.8-27b",
    "messages": [{"role": "user", "content": "Write a haiku about GPUs."}]
  }'

from openai import OpenAI
import os

client = OpenAI(base_url="https://osarai.xyz/v1", api_key=os.environ["OSARAI_API_KEY"])
reply = client.chat.completions.create(
    model="qwen3.8-27b",
    messages=[{"role": "user", "content": "Write a haiku about GPUs."}],
)
print(reply.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://osarai.xyz/v1", apiKey: process.env.OSARAI_API_KEY });
const reply = await client.chat.completions.create({
  model: "qwen3.8-27b",
  messages: [{ role: "user", content: "Write a haiku about GPUs." }],
});
console.log(reply.choices[0].message.content);