Connect any LLM
Settings → Language model. Keys are encrypted on your machine.
| Provider | Base URL | Example model |
|---|---|---|
| openai | https://api.openai.com/v1 | gpt-4.1-mini |
| anthropic | https://api.anthropic.com/v1 | claude-sonnet-5-5 |
| gemini | https://generativelanguage.googleapis.com/v1beta/openai | gemini-2.5-flash |
| groq | https://api.groq.com/openai/v1 | llama-3.3-70b-versatile |
| openrouter | https://openrouter.ai/api/v1 | openai/gpt-4.1-mini |
| deepseek | https://api.deepseek.com/v1 | deepseek-chat |
| ollama | http://localhost:11434/v1 | llama3.1 |
| custom | your URL | – |
Choose custom for any OpenAI-compatible endpoint, such as LM Studio, vLLM, Together, Mistral or Azure OpenAI. For a fully local setup, install Ollama and pick ollama.
Image model
The Media studio uses any OpenAI-compatible Images API, for example gpt-image-1 or dall-e-3. Set it in Settings → Language model → Image model.
From Claude or another agent
When an agent drives LinkedIn Agent over MCP, it can write the text itself with get_skill_prompt and save_draft. No model key is needed in that case.