Settings & LLM keys
Open Settings with the gear icon in the desktop header.
AI / LLM
Choose one provider profile:
- Ollama — Ollama Cloud / API models via your Ollama account, or a local Ollama server if you run models on your PC
- OpenRouter — API key and credits
- Custom — OpenAI-compatible base URL and key (including
https://ollama.com/v1or a local endpoint)
Pick the active model in the header after the provider can list or accept models. Keys are stored locally (encrypted via the local LLM gateway) — not in browser localStorage on the marketing site.
Cloud vs local inference
- Cloud (Ollama Cloud, OpenRouter, hosted custom APIs): inference runs on the provider’s servers. It does not drive Locimens RAM beyond the Locimens + MT5 baseline.
- Local models: PC requirements depend on the model you load. Add that model’s published CPU/RAM/disk (and optional GPU) needs on top of Locimens + MT5. See Download & install.
Widgets
Toggle floating panels: P&L, bulk close, open positions, price watch, and agent chat. Optional grouping locks stacked panels.
Telegram
Optional bot token and chat ID for remote notifications and commands. See Telegram bot.
MCP
Bridge URL(s), discovery, and health checks. The bridge must be reachable for tools and plugin UIs.
MT5
Mailbox path and Experts-related options for EA integration.
Privacy tip
Rotate API keys if you share logs or screen recordings.