The Archive · CLI · Guides

Set up a provider and a model

Run the guided setup, give TALOS a key or sign in, choose a model, and check that it answers.

Checked on CLI 0.3.3

A newer version is out (CLI 0.5.2): some details may differ.

Install the CLI first: Install the CLI.

The guided setup

Open a terminal in a project and run talos.
The first time, a short setup asks only what is missing, one step at a time:
- whether to trust the folder;
- a provider — a local engine that is running comes first;
- its key, typed in hidden input;
- a model.
Esc skips a step. /setup in the interactive screen, or talos setup from the shell, runs it again.

A key already in your environment, such as OPENAI_API_KEY, is used only after you agree once. A key you type is kept in the operating system’s credential store, never in a file.

From the shell

talos provider set-key openai          # asks for the key without echoing it, tests it, saves it
talos provider login openrouter        # or sign in to OpenRouter in the browser
talos model use openai:gpt-5-mini      # choose the model (and so its provider)
talos provider test openai             # check that it answers, without generating tokens

A local engine — Ollama, LM Studio or llama.cpp — needs no key. To point TALOS at one that is not on its default address, use talos provider configure <provider> --endpoint <url>.

In the interactive screen

/provider sets up a provider.
/model, or Alt+P, picks the model; with no provider chosen yet, the list of providers comes first.
/effort sets the reasoning effort, for models that support it.

If it does not answer

talos provider test <provider> prints its state, latency and what to do; talos doctor says what is ready and what is missing. See Check your setup and report a problem.