Quickstart

Call a shared Model API endpoint. Baseten hosts the model and manages the infrastructure; you send requests without deploying a model. You pay for the input and output tokens your requests use. See pricing and limits.

To have your coding agent run this quickstart, see Run a quickstart with your agent.

Create an API key

Use a personal API key to authenticate your requests. Reuse a saved key or create one:

Open API keys and sign in. If you don’t have an account, create one first.
Select Create API key and enter a name such as quickstart.
Select Personal, then Create API key.
Copy the key. Baseten only shows it once.

In your terminal, replace YOUR_API_KEY with your key and run:

export BASETEN_API_KEY="YOUR_API_KEY"

$env:BASETEN_API_KEY = "YOUR_API_KEY"

Run the examples from this terminal so they can read BASETEN_API_KEY. Keep the key out of source code and version control.

Send a request

Call the Chat Completions API with Python, JavaScript, or cURL. The SDK examples use Baseten’s base URL and API key; you don’t need an OpenAI account. The model field selects the hosted model, GLM 5.3 in this example, and messages contains your input.

Python
JavaScript
cURL

You’ll need uv to run this example.Save this code as chat.py:

chat.py

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://inference.baseten.co/v1",
    api_key=os.environ["BASETEN_API_KEY"],
)

response = client.chat.completions.create(
    model="zai-org/GLM-5.3",
    messages=[
        {"role": "user", "content": "Explain inference in one sentence."}
    ],
)

print(response.choices[0].message.content)

Run it. --with openai installs the SDK for this run:

uv run --with openai chat.py

Inference uses a trained model to generate outputs from new inputs.

You’ll need Node.js. Install the OpenAI SDK in your project directory:

Terminal

npm install openai

Save this code as chat.mjs:

chat.mjs

import OpenAI from "openai";

const client = new OpenAI({
    baseURL: "https://inference.baseten.co/v1",
    apiKey: process.env.BASETEN_API_KEY,
});

const response = await client.chat.completions.create({
    model: "zai-org/GLM-5.3",
    messages: [
        { role: "user", content: "Explain inference in one sentence." }
    ],
});

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

Run it:

node chat.mjs

Inference uses a trained model to generate outputs from new inputs.

Run this command in a Bash-compatible terminal:

curl https://inference.baseten.co/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $BASETEN_API_KEY" \
  -d '{
    "model": "zai-org/GLM-5.3",
    "messages": [
      {"role": "user", "content": "Explain inference in one sentence."}
    ]
  }'

{
  "choices": [
    {
      "message": {
        "role": "assistant",
        "content": "Inference uses a trained model to generate outputs from new inputs."
      }
    }
  ]
}

The reply is in choices[0].message.content.

The exact wording varies. Change the message and run the request again.

Next steps

Choose another Model API and replace the model value with its slug.
Stream the response to display text as the model generates it.
Deploy from Explore if you need a model on dedicated GPUs with your own scaling settings.

Get started

Model APIs

Inference

Training

Baseten Grid

Frontier Gateway

Sandboxes

Development

Deployment

Engines

Model Labs

Organization

Hosting options

Observability

Troubleshooting

Create an API key

Send a request

Next steps

​Create an API key

​Send a request

​Next steps

Create an API key

Send a request

Next steps