Getting Started

Give your local AI persistent memory in under 5 minutes.

1. Install

pip install velqua

Requires Python 3.10+. Works on Linux, macOS, and Windows.

2. Start the server

velqua-server

Opens http://localhost:8765. The setup wizard walks you through everything.

3. Choose your provider

Ollama — local models, zero config
OpenAI — enter your API key
Anthropic — enter your API key
Groq — enter your API key

4. Import your history

Export from ChatGPT (Settings → Data Controls → Export) or Claude (Settings → Export). Drag the JSON into the Import tab. Velqua auto-detects the format. From 1,000 conversations, typically 50–200 high-quality facts are extracted.

5. Point your apps at the proxy

Before: localhost:11434
After:  localhost:11435

That's it.

Troubleshooting

Model says it doesn't know me

Make sure you imported at least one conversation export. Check the Facts tab. If the count is 0, re-import.

Port 11435 is already in use

Change the proxy port in Settings. Any available port works. Update your apps accordingly.