AI Coding

Continue CLI in Russia: Installation, config.yaml, and API Setup

Install Continue CLI, select the active config.yaml, add an OpenAI-compatible model safely, and verify interactive and headless requests.

BetterToken

Editorial Team

Aug 5, 20264 min read

Continue CLI installs as the cn command and reads the same YAML configuration schema as Continue in the editor. An OpenAI-compatible connection needs the openai provider, a Base URL, the current Model ID, and a safe reference to the API key. Verify installation with cn --version, then run one short interactive or headless request.

Want to test Continue CLI with your own API key without storing the secret in the repository? Create a BetterToken account and keep the current Continue CLI guide open. The Model ID is deliberately a placeholder in the example; copy the current value on the day you configure it.

What you need

a terminal that can reach the official Continue installer;
your own BetterToken API key;
the current Model ID;
a user directory for a local config.yaml;
a short test prompt with no secrets or access to important files.

BetterToken covers the OpenAI-compatible API endpoint. It does not guarantee access to GitHub, npm, the Continue website, the installer, or updates. The BetterToken endpoint does not require a VPN from Russia; verify the other network steps separately.

1. Install Continue CLI

The current Continue documentation provides a shell installer for macOS/Linux:


curl -fsSL https://raw.githubusercontent.com/continuedev/continue/main/extensions/cli/scripts/install.sh | bash

On Windows, use the current PowerShell command from the official documentation. If you prefer npm and have a supported Node.js version installed:


npm install -g @continuedev/cli

Do not run both methods in sequence after the first error. First determine which binary and directory were already created.

Verify the installation:


cn --version

The command should print a version without command not found. This article does not pin a version number.

2. Find the active config.yaml

Continue CLI selects configuration in this order:

the path supplied with --config <path>;
the last saved configuration;
the default user file.

Default path on macOS/Linux:


~/.continue/config.yaml

On Windows:


%USERPROFILE%\.continue\config.yaml

If changes are not applied, specify the file explicitly with --config instead of editing several copies at random.

3. Add an OpenAI-compatible model

Minimal YAML:


name: BetterToken Continue CLI
version: 0.0.1
schema: v1

models:
  - name: BetterToken model
    provider: openai
    model: YOUR_MODEL_ID
    apiBase: https://www.bettertoken.ai/v1
    apiKey: ${{ secrets.BETTERTOKEN_API_KEY }}

Replace YOUR_MODEL_ID with the exact current ID. Secret-reference syntax depends on the current Continue version. If the CLI uses a process variable, export it in the local session and follow the current configuration page.

Never store a real key in the repository. If config.yaml is shared, keep only the secret reference and add the local credentials file to the tool's ignore configuration.

4. Run an interactive test

Start the CLI with an explicit configuration to rule out a precedence problem:


cn --config ~/.continue/config.yaml

Send a short request:


Reply with CONTINUE_OK and do not change files.

If the first run displays onboarding, confirm that the CLI can see the local configuration. Do not put the key in the prompt or a shared log to bypass the screen.

5. Run a headless test

For a one-shot check, use the current cn -p syntax:


cn --config ~/.continue/config.yaml -p "Reply with CONTINUE_OK and do not change files."

The test succeeds when the command completes with a normal response, uses the expected Model ID, and creates a BetterToken Dashboard record with the model, status, and token charge.

Common errors

`cn: command not found`

Restart the shell and check PATH. If you installed with npm, locate the global binary directory. Do not install a second copy before checking.

CLI ignores config.yaml

Pass --config with an absolute path. After a successful test, determine which persisted or default configuration had been selected before.

`401`

Check the secret reference, whether the variable exists in the current process, and spaces in the key. Do not print the variable value.

`404`

For the OpenAI-compatible provider, the Base URL must be:


https://www.bettertoken.ai/v1

Do not append a complete endpoint path when apiBase expects the API root.

`model not found`

Copy the current Model ID from the BetterToken catalog. Check capitalization and suffixes; do not use a name from an old article.

What the test actually confirms

The short request confirms that the current cn read the intended YAML, obtained the secret, contacted the correct endpoint, and used an available Model ID. It does not confirm every tool, remote session, cloud feature, or Continue integration. Test those separately with safe, bounded requests.

Ready to optimize your LLM workflow?

Join thousands of developers building faster, smarter, and more cost-effective AI applications with BetterToken.