PacketLab API
Connect existing tools to PacketLab in three steps.
Top up, create an API key, then replace Base URL and use modelpass/auto. Model access, point billing, and usage logs live in one console.
1. Use Auto
Smart-routing clients only use modelpass/auto; the platform automatically selects the right capability.
2. Top up and create key
Top up points, create an API key, and store it as a server-side environment variable.
3. Replace Base URL
Keep the existing SDK and replace only baseURL, apiKey, and model.
curl "https://packetlab.app/v1/chat/completions" \
-H "Authorization: Bearer $PACKETLAB_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "modelpass/auto",
"max_tokens": 800,
"temperature": 0.7,
"messages": [
{ "role": "user", "content": "Write a short product brief." }
]
}'import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.PACKETLAB_API_KEY,
baseURL: "https://packetlab.app/v1",
});
const completion = await client.chat.completions.create({
model: "modelpass/auto",
max_tokens: 800,
temperature: 0.7,
messages: [{ role: "user", content: "Write a short product brief." }],
});
console.log(completion.choices[0].message.content);Billing rules
Each model consumes points by public pricing; only completed responses are billed.
Compatible parameters
Supports model, messages, max_tokens, temperature, top_p, stop, response_format, frequency_penalty, presence_penalty, and seed. stream=true is not supported yet.
Technical advantages
One API entry point, one wallet, and one usage ledger make multi-model access easier for products and developer tools.