Setup

Open the Agentic Chat settings tab after installing the plugin. The main setup choice is the model provider.

OpenRouter

OpenRouter is the default hosted provider path.

1. Create an API key at openrouter.ai/keys.
2. Paste it into Settings > Agentic Chat > Models.
3. Pick a tool-capable model. The default model is moonshotai/kimi-k2.6.
4. Keep the strict privacy routing defaults enabled unless you deliberately want a broader provider set.

With the default privacy settings, model browsing only shows options that can satisfy the zero-data-retention routing requirements.

Ollama

Use Ollama when you want fully local model calls.

1. Run a local Ollama server.
2. Pull a tool-capable model.
3. Set the provider to Ollama (local).
4. Keep the default server URL, http://localhost:11434, unless your Ollama server is elsewhere.
5. Enter the Ollama model tag.

Ollama requests stay on your device unless your own Ollama setup routes them elsewhere.

OpenAI-compatible gateways

Choose OpenAI-compatible for gateways such as OpenWebUI, LM Studio, vLLM, llama.cpp, Chutes, or Venice.ai.

Set the base URL to the gateway root whose /chat/completions endpoint is valid. For OpenWebUI you can use the site root or the API base directly; the plugin resolves bare OpenWebUI roots to /api. The default local OpenWebUI API base is:

http://localhost:3000/api

Then paste the gateway bearer token and model id exposed by that gateway.

Thinking level

Some models support adjustable reasoning effort. Set it in Settings > Agent > Thinking level:

Level	Behavior
off	No extra reasoning tokens.
minimal	Light internal reasoning.
low	Moderate reasoning.
medium	Balanced depth.
high	Deep reasoning, more tokens.
xhigh	Maximum reasoning depth.

You can also set effort per message with /effort [level] in chat.

Proxy settings

On desktop, Models > Network proxy > HTTP proxy lets plugin-owned model, model-browsing, web, MCP, and observability requests use an HTTP proxy such as:

http://host:port

On mobile, leave plugin proxy fields empty and use the device, VPN, or network-level proxy path.

Semantic retrieval setup

If you want vector-based note search, enable Settings > Resources > Semantic retrieval:

1. Choose an embedding provider (OpenRouter, Ollama, or OpenAI-compatible).
2. Enter the embedding model id.
3. Set vector dimensions to match the model (default 1536).
4. Use Ollama for fully local embeddings if you do not want note content sent to a remote provider.

After setup, run /semantic-index start in chat to build the index for your current scope.