Quickstart

From a Hugging Face link to a working API call, in five steps.

1. Launch a model

Sign in with your litepod account, open the dashboard and paste a Hugging Face model link, for example https://huggingface.co/Qwen/Qwen3-4B. The model must be public (not gated), at most 10B parameters, and must not need custom code. Pilot checks it, picks a GPU and deploys it. See the models list for ones that work well.

2. Get the model id

When the launch is ready, the dashboard shows the model id, which looks like modeldock/<model>-<suffix>. You can also list your models with GET https://api.lab.litepod.io/v1/models.

3. Create an API key

Create a key in the dashboard. It starts with sk_modeldock_. Keep it secret; see Authentication.

4. Make your first call

curl https://api.lab.litepod.io/v1/chat/completions \
  -H "Authorization: Bearer $MODELDOCK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "modeldock/<your-model-id>",
    "messages": [{"role": "user", "content": "Say hello in five words."}]
  }'

from openai import OpenAI

client = OpenAI(base_url="https://api.lab.litepod.io/v1", api_key="sk_modeldock_...")

reply = client.chat.completions.create(
    model="modeldock/<your-model-id>",
    messages=[{"role": "user", "content": "Say hello in five words."}],
)
print(reply.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.lab.litepod.io/v1", apiKey: process.env.MODELDOCK_API_KEY });

const reply = await client.chat.completions.create({
  model: "modeldock/<your-model-id>",
  messages: [{ role: "user", content: "Say hello in five words." }],
});
console.log(reply.choices[0].message.content);

5. What happens after the test

A Pre-flight ends after 15 minutes or 25 requests. Calls to an ended Pre-flight return the trial_ended error. Launch again from the dashboard to keep testing; see Errors and limits for how many free Pre-flights an account gets.

More from litepod: the free LLM API with free open-source models. ModelDock is a litepod product.