Ruthless.

Novice · Coding tools · 7 min read

Connect Ruthless to your coding editor, one setting at a time

Understand the key, base URL and model ID before configuring Continue or Cline, then distinguish a text connection from working coding tools.

← All guidesNovice guidesBy Ruthless AI team

Learn by using GetRuthless

Keep this guide beside your Ruthless chat. Copy one prompt, add the example it asks for and paste it into “Message Ruthless”. Nothing is sent until you choose to send it.

Open Ruthless in another tab ↗More in Ruthless in your coding toolsNew to this? Start here

The short answer

Generate a Ruthless API key privately, use https://getruthless.io/v1 as the base URL and choose an exact available model ID. Configure an OpenAI-compatible client to use Chat Completions. Keep Continue’s useResponsesApi set to false. A text response checks basic chat; editing and tool use need their own small checks.

Understand the three pieces you will enter

An IDE is a coding editor: software for reading files, changing them and running your project. Continue and Cline are extensions that add AI assistance to such an editor. The API is the agreed format through which the extension asks Ruthless for a response. This connection does not make the hosted website chat see your local files.

The API key is your private credential. It identifies your workspace for requests and spending, so enter it only into the extension’s private credential setting. The base URL is the service address: https://getruthless.io/v1. The model ID is the exact machine-readable name of the model you want, rather than its friendly display name. Check the current catalogue and choose for your coding task; kimi-pro-abliterated is the specific example used here, not a claim that an abliterated model is best for coding.

Open GetRuthless and use Generate API key. Keep the value out of chat, screenshots and shared code. Public GET /v1/models, at https://getruthless.io/v1/models, discovers available IDs; it does not validate your key. Ruthless API requests use your shared workspace wallet. Check /pricing and Usage & balance before deciding to send paid requests.

Set up Continue without choosing the wrong protocol

Open Continue in your editor and locate its model configuration. Use the OpenAI-compatible provider fields documented by Continue: provider, model, apiBase and apiKey. [1] “OpenAI-compatible” means the request follows a familiar format; you are configuring Ruthless as the service. The field useResponsesApi must remain false because this setup uses Chat Completions, a particular request-and-answer format. Compatibility does not imply support for the different Responses API.

If you are unsure where to enter settings, use the teaching prompt below in GetRuthless chat first. Supply your extension version and the field names you see, with no key value. Ask for instructions matching those controls instead of pasting an unfamiliar configuration into your project.

Try this prompt
Teach me to configure Continue for Ruthless. I understand that the key is private and will enter it only in the extension. Base URL: https://getruthless.io/v1. Example model ID: kimi-pro-abliterated, which I will check in the current catalogue. Explain provider: openai, apiBase and useResponsesApi: false in beginner language. Tell me what non-secret information you need about my Continue version. Do not request my credential or claim you have tested my editor.

Use Cline’s corresponding connection fields

In Cline’s provider settings choose OpenAI Compatible. Enter the Ruthless base URL, your key privately and an exact available model ID in the corresponding fields. Cline documents this provider option. [2] Consult Ruthless API docs for supported request behaviour. [3] These are configuration directions, not evidence that your installed extension and selected model have passed an integration test.

Keep your first task to a plain text question. If you choose to make a connection check, a tiny request such as “Reply CONNECTION_OK only” still uses the shared wallet at applicable rates. A received answer and the workspace usage entry support basic authentication and text chat. They do not establish that file reading, streaming or tool calling work.

Judge failures before repeating requests

If connection fails, note the extension version, selected provider, model ID and a redacted error. Ask Ruthless to distinguish a wrong address, unavailable model and unsupported request format. Never include an Authorization header, which can carry the credential. If a request times out, inspect the client’s status and Usage & balance before retrying; a missing answer does not prove nothing was processed.

Your exercise is to explain the three settings back in your own words and save a non-secret setup note. Then, when ready, try explaining one harmless local file before permitting edits. The extension controls which files it sends and which tools it runs; inspect those permissions separately.

Common questions

Does the public model list prove my key works?

No. It is discovery. An authenticated request is a different operation and may incur usage charges.

Should I paste my key into Ruthless to troubleshoot?

No. Share field names, extension version and a redacted error. Keep the credential in private client storage.

Sources & further reading

  1. Continue: OpenAI-compatible provider configuration
  2. Cline: OpenAI Compatible provider
  3. Ruthless API documentation

The example prompts and workflows are GetRuthless guidance. Source references support the concepts explained; they are not an endorsement of GetRuthless.

Put the guide into practice.

Open Ruthless and choose an available standard or abliterated model suited to your task. Copy a prompt above, add the context it asks for and check the result. You can compare models in the same workspace; see current pricing before sending requests.