Use OpenRouter with Cline

Cline’s OpenRouter page is the setup source. OpenRouter’s own docs index fetched the same day did not include a Cline guide (a guessed coding-agents URL returned 404). Pricing sentences below are labeled by which site said them.

Key

• Cline says to sign in at openrouter.ai with a Google or GitHub account.
• Open the keys page. Cline says an API key may already be listed; if not, create one and copy it.

Cline settings

• Click the settings icon in the Cline panel.
• Choose “OpenRouter” from the API Provider dropdown.
• Paste the key into the “OpenRouter API Key” field.
• Choose a model from the Model dropdown.
• Custom base URL is optional. Cline says to leave it blank for most users.

Caching and price

Cline says OpenRouter charges based on the underlying model’s pricing, and that caching should activate automatically when the model supports it. The exception Cline documents is Gemini models via OpenRouter: you must check “Enable Prompt Caching” in the provider settings. Cline says that checkbox is not necessary for non-Gemini models on OpenRouter.

OpenRouter’s FAQ, separately, says inference is passed through with no markup, and that a fee is charged when you purchase credits. The FAQ markdown leaves the card-fee formula unexpanded, so this page does not state a percent. Cline’s page does not mention that credit-purchase fee. Free-model request caps, for ids ending in :free, are on the OpenRouter limits page: 20 requests per minute, and 50 or 1,000 requests per day.

Other notes

• Use Ollama local models with OpenCode
• Use Groq with Cline and OpenCode
• Prices the docs attach to coding models
• Token price calculator
• LLM APIs category list
• Free LLM API limits