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

Run the agent with Ollama locally — no API keys, $0

Run remoteagent.online fully offline with Ollama. No API keys, prompts never leave your machine, and the model cost is zero.

Updated 2026-08-22Reading time 4 min

Why Ollama

Ollama runs models locally. You pay nothing per token and no key ever leaves your machine. It is the default choice when the task must stay private.

Install Ollama

  1. Install Ollama from ollama.com.
  2. Pull a model: ollama pull llama3.1.
  3. Confirm it runs: ollama run llama3.1 "hi".

Point the agent at it

  1. In the console, open Settings → Provider.
  2. Choose Ollama and set the local endpoint (usually http://localhost:11434).
  3. Pick the model you pulled.
  4. Save — no API key required.

Test it

$ remoteagent chat "what are you running on?"

Trade-offs

i
Local models are free and private but slower and less capable than frontier models. Use Ollama for private or routine work, and a provider key for the hard tasks.