Quick Start

From zero to a first answer in three steps. The examples use the official OpenAI SDKs, plain cURL or the railwail npm package; pick your language in the tabs.

1. 1

   Create an API key

   Sign in and open API keys. Keys start with rw_live_ and are shown once, so copy yours right away.

   A new key has the scopes read and chat. That is enough for this page. For images, video, audio or embeddings tick those scopes (or all) when you create the key; otherwise those calls return 403 insufficient_scope.

2. 2

   Put the key in your environment

   Every example in these docs reads the key from RAILWAIL_API_KEY, so they run as copied.

   bashPowerShell.env

   Copy

   export RAILWAIL_API_KEY="rw_live_..."

   Check the key for free: listing one model costs nothing and answers 401 if the key is wrong.

   curl -s "https://railwail.com/api/v1/models?limit=1" \
     -H "Authorization: Bearer $RAILWAIL_API_KEY"

3. 3

   Make your first request

   Chat with gpt-4o-mini. max_tokens caps the answer, and with it the credits the API reserves for the call.

   cURLPython (OpenAI SDK)Node (OpenAI SDK)railwail SDK

   Copy

   // npm i openai   (ESM: save as quickstart.mjs)
   import OpenAI from "openai";
   
   const client = new OpenAI({
     apiKey: process.env.RAILWAIL_API_KEY,
     baseURL: "https://railwail.com/api/v1",
   });
   
   const completion = await client.chat.completions.create({
     model: "gpt-4o-mini",
     messages: [{ role: "user", content: "Explain quantum computing in one paragraph." }],
     max_tokens: 300,
   });
   console.log(completion.choices[0].message.content);

More examples#

Chat with a system prompt and history

Pass the whole conversation; the model sees every message.

curl https://railwail.com/api/v1/chat/completions \
  -H "Authorization: Bearer $RAILWAIL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 400,
    "messages": [
      {"role": "system", "content": "You are a helpful coding assistant."},
      {"role": "user", "content": "How do I reverse a string in Python?"}
    ]
  }'

completion = client.chat.completions.create(
    model="claude-sonnet-4-6",
    max_tokens=400,
    messages=[
        {"role": "system", "content": "You are a helpful coding assistant."},
        {"role": "user", "content": "How do I reverse a string in Python?"},
    ],
)
print(completion.choices[0].message.content)

const completion = await client.chat.completions.create({
  model: "claude-sonnet-4-6",
  max_tokens: 400,
  messages: [
    { role: "system", content: "You are a helpful coding assistant." },
    { role: "user", content: "How do I reverse a string in Python?" },
  ],
});
console.log(completion.choices[0].message.content);

const reply = await rw.run(
  "claude-sonnet-4-6",
  [
    { role: "system", content: "You are a helpful coding assistant." },
    { role: "user", content: "How do I reverse a string in Python?" },
  ],
  { max_tokens: 400 },
);
console.log(reply);

Generate an image

Needs a key with the images scope. The response carries a URL per image; it can be a temporary provider link, so download the file if you want to keep it.

curl https://railwail.com/api/v1/images/generations \
  -H "Authorization: Bearer $RAILWAIL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "flux-1-schnell",
    "prompt": "A cyberpunk cityscape at dusk, neon reflections on wet streets"
  }'

image = client.images.generate(
    model="flux-1-schnell",
    prompt="A cyberpunk cityscape at dusk, neon reflections on wet streets",
)
print(image.data[0].url)  # None while a slow run is still processing (HTTP 202)

const image = await client.images.generate({
  model: "flux-1-schnell",
  prompt: "A cyberpunk cityscape at dusk, neon reflections on wet streets",
});
console.log(image.data[0].url); // null while a slow run is still processing (HTTP 202)

const res = await rw.image(
  "flux-1-schnell",
  "A cyberpunk cityscape at dusk, neon reflections on wet streets",
);
console.log(res.data[0].url);

Embeddings

Vectors for search and retrieval, with the same key (it needs the embeddings scope). Limits and errors are on the embeddings page.

res = client.embeddings.create(model="text-embedding-3-small", input=["first text", "second text"])
print(len(res.data), len(res.data[0].embedding))  # 2 1536

const res = await client.embeddings.create({ model: "text-embedding-3-small", input: ["first text", "second text"] });
console.log(res.data.length, res.data[0].embedding.length); // 2 1536

Next steps#