Skip to content
← Help center

Getting started

Make your first Stimulir call and find it in Traces

Production guidance · Console v0.2 · Updated

Before you start

Stimulir is OpenAI-compatible. You need a project in the Console and its project ID. Select the project in the Console and copy the ID.

Install the CLI and create a key

  • Follow CLI installation to set up the command for your machine. That guide covers uv, pip, updates and PATH troubleshooting.
  • Run stimulir --version. These instructions were checked against the published 0.2.2 CLI; upgrade an older installation before using newer commands.
  • Run stimulir login.
  • Run stimulir keys create --name first-call --save. The key begins with hyb_ and is shown once, so copy it.
  • Set STIMULIR_API_KEY to the key and STIMULIR_PROJECT_ID to your project ID in your environment. Do not paste real keys into support messages, screenshots, or source control.
  • Run stimulir models and pick a model that is enabled for your project.

Send the request

Send a chat completion to https://api.stimulir.com/api/v1/inference/chat/completions. Use your key as the bearer token and send your project ID in the X-Project-Id header. With the OpenAI SDK, set the base URL to https://api.stimulir.com/api/v1/inference and pass the project header as a default header. The developer quick start has complete Python examples; the CLI quickstart includes curl. Include a stable workflow tag so you can find the request.

Check the result in Traces

After a successful call, open the Console, select the same project, and open Traces. Check the model, tokens, latency, and cost where they are available. Confirm that actual input/output was captured before reusing the trace in an evaluation; a metering placeholder alone is not usable conversation evidence.

Other ways to start

Sources and scope

Reviewed against a verified production release. The release record and source-code evidence are retained internally. Check the Console version and workspace availability in your environment.