Build on ApiSoul

Create a workspace key in the console, then point existing OpenAI SDK code at https://apisoul.store/api/v1. Public shortcut routes are also available under https://apisoul.store/v1 when the front proxy forwards them.

Quickstart

npm install openai

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.APISOUL_API_KEY,
  baseURL: "https://apisoul.store/api/v1",
});

Authentication

Send your workspace API key as a bearer token. Raw keys are shown once and stored only as hashes.

Authorization: Bearer sk_live_rg_...

API keys

Open the dashboard, verify your email, then create a key with a route group, rate limit, optional monthly credit limit, and optional expiry date.

Chat completions

const result = await client.chat.completions.create({
  model: "gpt-5.5",
  messages: [{ role: "user", content: "Write a deployment checklist." }],
  max_tokens: 500
});

console.log(result.choices[0].message.content);

Streaming

const stream = await client.chat.completions.create({
  model: "gpt-5.5",
  messages,
  stream: true,
  stream_options: { include_usage: true }
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content || "");
}

For accurate streamed billing, request an upstream usage chunk with stream_options.include_usage. If upstream omits usage, ApiSoul conservatively charges the reserved estimate.

Models

const models = await client.models.list();
console.log(models.data.map(model => model.id));

Errors

{
  "error": {
    "type": "billing_error",
    "code": "insufficient_balance",
    "message": "Insufficient credit balance.",
    "request_id": "req_..."
  }
}

Metering

Each request creates a reservation, then reconciles actual prompt and completion tokens after the provider response. Failed requests release the reservation when no provider usage is recorded.