just let it

DISTRIBUTED AI ORCHESTRATION

// OVERVIEW

CLUSTER YOUR FLOCK. SATURATE YOUR HARDWARE.

ClusterFlock unifies your heterogeneous GPU fleet - NVIDIA, Apple Silicon, DGX, consumer cards - into a single AI backend. Load models, run inference, and launch autonomous missions across every device from one command.

// CAPABILITIES

WHAT IT DOES

⎔

SMART ALLOCATION

Auto-detects VRAM, profiles hardware, and bin-packs the best models onto each GPU. No manual config. Self adapting and self healing, even while the mission runs.

⟁

MIXTURE OF AGENTS

A showrunner LLM coordinates a flock of worker models - dispatching tasks, evaluating results, iterating until done.

◈

AUTONOMOUS MISSIONS

Describe a goal. ClusterFlock spins up sandboxed containers, assigns agents, and builds the solution autonomously.

⟟

REAL-TIME TELEMETRY

Live GPU utilization, VRAM, model status, and tokens/sec from every node.

⊞

MULTI-BACKEND

Native support for llama.cpp, LM Studio, Metal, and CUDA. DGX Spark, consumer GPUs, and Mac - all in one cluster.

⎊

OPENAI-COMPATIBLE API

Drop-in replacement. Point any OpenAI SDK, LangChain app, or curl command at port 1919 and go. ClusterFlock intelligently routes across your network and saturates your AI hardware with work.

// API

ONE ENDPOINT. FULL CLUSTER.

POST http://your-cluster:1919/v1/chat/completions { "model": "clusterflock", "messages": [ {"role": "user", "content": "Explain quantum computing"} ] }

Three routing modes: FANOUT broadcast to all, synthesize best SPEED fastest single endpoint MANUAL pick your model

Works with OpenAI SDK · LangChain · LiteLLM · curl

// OPEN SOURCE (MIT)

AVAILABLE SOON!

We're putting in some final touches.
- your friends at Notum Robotics