Quickstart

This page takes you from no account to a working request. If you already have a key, skip to step three.

1. Request access

   XField is invite-only. Request access and tell us briefly what you plan to use it for. Pricing is shared with your invitation.

2. Sign in and create a key

   Once you are invited, sign in and open Settings › API keys. Create a key and give it a name you will recognize later, such as the machine or project that uses it.

   Keys start with sk-xf-v1-. The full key is shown once, so copy it before you close the dialog.

3. Export the key

   Keep the key in an environment variable rather than in your code.

   ShellCopy

   export XFIELD_API_KEY="sk-xf-v1-..."

   Add the line to your shell profile or your project's .env file if you want it to persist.

4. Send your first request

   XField speaks the OpenAI API, so the official SDKs work as they are. Set the base URL and pick a model.

   PythonJavaScriptcURLCopy

   PythonCopy

   import os
   from openai import OpenAI
    
   client = OpenAI(
       base_url="https://xfield.cloud.brutal.ai/api/v1",
       api_key=os.environ["XFIELD_API_KEY"],
   )
    
   response = client.chat.completions.create(
       model="deepseek-v4.1-flash-1017-uncensored",
       messages=[
           {"role": "user", "content": "Explain how a format string vulnerability works."}
       ],
   )
    
   print(response.choices[0].message.content)

   JavaScriptCopy

   import OpenAI from "openai";
    
   const client = new OpenAI({
     baseURL: "https://xfield.cloud.brutal.ai/api/v1",
     apiKey: process.env.XFIELD_API_KEY,
   });
    
   const response = await client.chat.completions.create({
     model: "deepseek-v4.1-flash-1017-uncensored",
     messages: [
       { role: "user", content: "Explain how a format string vulnerability works." },
     ],
   });
    
   console.log(response.choices[0].message.content);

   ShellCopy

   curl https://xfield.cloud.brutal.ai/api/v1/chat/completions \
     -H "Authorization: Bearer $XFIELD_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{
       "model": "deepseek-v4.1-flash-1017-uncensored",
       "messages": [
         {"role": "user", "content": "Explain how a format string vulnerability works."}
       ]
     }'

   You should get a normal Chat Completions response back. DeepSeek V4.1 Flash thinks before it answers, so the response also carries a reasoning field. See Reasoning to tune or turn it off.

5. Go further

   You are set up. From here, most people do one of these:

   • Pick the right model for the job on the Models page.
   • Stream tokens as they arrive with Streaming.
   • Use the Anthropic SDK or the Responses API: see OpenAI and Anthropic APIs.
   • Run a coding agent against XField with the Claude Code setup.
   • Give the model functions to call with Tool calling.