Personal operations
Use local context to handle recurring work across files, email, browser, and reminders.
Local AI agent
A local-first AI agent that lives on your machine.
Your agent's identity, memory, vault, wallet, and working context stay on your own machine, with your keys and your data. You still route each task to the best model for the job, from local Ollama to the top cloud providers.
Common jobs
Developer workflows
Keep project knowledge next to the code and reuse it on every coding task.
Long-running goals
Let the background mind keep working between conversations and across restarts.
Your data stays with you
EloPhanto is built around local files, local memory, and an encrypted vault, not a hosted account you rent. It runs with no telemetry by default.
- A persistent knowledge base on disk you control
- Encrypted credential and wallet vault
- No telemetry by default
Local doesn't mean cut off
It still reaches the outside world when a job needs it: cloud or local models, real browser sessions, every messaging channel, and community skills.
- Seven LLM providers, plus local Ollama
- CLI, web, VS Code, Telegram, Discord, and Slack
- One-command skills from EloPhantoHub
Do I have to self-host it?
EloPhanto is local-first: it runs on your own machine, with your keys and your data. A managed cloud option is on the way for anyone who'd rather not self-host.
Can it use local models?
Yes. Run everything locally with Ollama, or route a task to a cloud model like GPT-5.5, Claude, or GLM when you want more power.