remoteagent.online  the open-source control plane for secure, cost-controlled AI agents

Run agents on local models with Ollama - zero token cost

Self-host Llama, Mistral, Qwen or Phi-3 and run RemoteAgent Online agents against them for zero per-token cost.

Updated 2026-08-24Reading time 5 min

Why local

Local models mean prompts never leave your machine, every token is free and offline work is possible.

Setup

Install Ollama, pull a model and point the agent at it:

ollama pull llama3.1
remoteagent model add local --ollama llama3.1
remoteagent model use local

Same policy, same audit

Local inference changes nothing else: policies, grants, caps and the audit log work identically.

When to use APIs

For frontier-quality reasoning, fall back to an API model per task - mix and match freely.

Next

Try a RAG agent on top of your local model.

RemoteAgent Online - 100 solutions · Blog · mona.expert · Sign in free