Client setup
OpenAI SDK setup
Use the OpenAI SDK when a Node.js service or script calls Lumoswitch directly. The API configuration must expose OpenAI-compatible output.
Install and configure
pnpm add openai
export LUMOSWITCH_BASE_URL="https://api.lumoswitch.com/v1"
export LUMOSWITCH_API_KEY="your-lumoswitch-access-key"
export LUMOSWITCH_MODEL="your-client-facing-model-name"Send a request
import OpenAI from "openai";
const client = new OpenAI({
baseURL: process.env.LUMOSWITCH_BASE_URL,
apiKey: process.env.LUMOSWITCH_API_KEY,
});
const response = await client.chat.completions.create({
model: process.env.LUMOSWITCH_MODEL,
messages: [{ role: "user", content: "Say hello in one sentence." }],
});
console.log(response.choices[0].message.content);The model value is the client-facing model name from the API configuration, not an upstream provider model ID.
Stream output
const stream = await client.chat.completions.create({
model: process.env.LUMOSWITCH_MODEL,
messages: [{ role: "user", content: "Write a short welcome message." }],
stream: true,
});
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}Troubleshooting
- For
401, verify thatapiKeyis a Lumoswitch access key. - For
404, ensurebaseURLends in/v1without/v1/v1. - If the model is missing, copy its client-facing model name from
/models. - Do not expose the access key in browser code; call Lumoswitch from your backend.