Local trading AI with Ollama
Point VaultCharts at Ollama or LM Studio for an on-network assistant. Chart tools still run against your vault; inference never needs a cloud LLM provider.
What you get
- ✓Ollama / LM Studio / OpenAI-compatible local endpoints
- ✓Same agentic tools as cloud providers
- ✓Maximum privacy for prompts and completions
How it works
1. Run Ollama locally
Pull a model, enable the local API, set base URL in VaultCharts Models settings.
2. Test connection
Confirm tags/models discover correctly before a live session.
Who it's for
- Traders who will not send prompts off-network
- Offline or air-gapped research setups
FAQ
Are local models as good as GPT/Claude?
Quality varies by model size. Tools still ground answers in your OHLC; weaker models may need clearer prompts.
Related
Next step
See the AI FAQ troubleshooting section for connection tips.