Every machine you own.
One local AI endpoint.

Run Ollama across multiple computers: Macs, Linux servers and Windows PCs, NVIDIA or not, as one endpoint for Claude Code, Codex and Open WebUI. Free and open source.

A home office with two Mac Studios and a tower PC by a window, captioned: You already own an AI cluster.
See it route in 60 seconds. Nothing loads from YouTube until you press play. What the demo shows
Ollama Herd dashboard showing two nodes side by side: a 512 GB Mac Studio running gpt-oss:120b and Qwen3-Coder models via Ollama and MLX, plus image generation (FLUX, SD3.5), speech-to-text, embeddings, and vision models; and a 128 GB MacBook Pro M4 running Gemma and Qwen3-Coder models.
The whole fleet on one page: every machine, every model loaded, every service ready to route. Linux and Windows nodes show up as the same kind of card.

Install it and your fleet joins the leaderboard within a day.

One URL in front. The right machine behind it.

Every request goes whole to the machine that already has its model loaded and room to run it. Nodes find the router on your network by themselves.

Claude Code, Codex, Open WebUIor any OpenAI, Ollama or Anthropic client
one endpoint
Herd routerdrops machines that can't serve the request, scores the rest on 8 signals
each request goes to one machine
Mac Studio70B model, loaded
Linux + RTXcoding model
Windows PCembeddings
Routing, not sharding: more machines serve more requests at once, and each model runs on one machine. How the scoring works · Routing or sharding?

Everything you run, on everything you own.

Two commands. No config files.

# Install (or upgrade to v0.12.5), works on macOS, Linux, and Windows
pip install ollama-herd --upgrade

# Start the router (on your most powerful machine)
herd

# On each machine running Ollama (Mac, Linux, or Windows)
herd-node

# That's it. Nodes auto-discover the router via mDNS.

No Kubernetes. No Docker. No YAML. Homebrew, uv and every install option →

Want one model split across several machines instead? That's sharding, what exo does. Which one do you need?

Your hardware deserves
an orchestrator

Stop leaving compute on the table. Start herding.