Zenthrix

Hardware-Adaptive Edge Neural Graph Compiler

Zenthrix is an edge-native model compiler frontend for compiling open-weight neural networks (LLMs, SLMs, and vision models) into zero-copy, memory-optimized binaries tailored for consumer edge silicon — Apple Silicon, Qualcomm Snapdragon NPU, and Arm Cortex/Ethos.

---

Table of Contents

• Key Features
• Installation
• Quickstart
• Supported Target Architectures
• Changelog
• Contributing
• License

---

Key Features

• Direct Ingestion — Native loaders for ONNX, PyTorch Export (AOTInductor), and GGUF architectures, with no intermediate format conversion required.
• Unified Memory Tiling — Schedules compute passes against unified memory architectures, reducing peak active RAM allocation by up to 40%.
• Zero-Copy Runtime — Emits standalone, relocatable .zx binaries that execute locally without a heavy Python runtime dependency.
• Privacy-First Compilation — Models compile entirely on-device; weights and computational graphs never leave the local environment.

Installation

Install the precompiled command-line client and runtime via pip:

pip install zenthrix

You can also install and run it with uv:

uv tool install zenthrix
zenthrix --version

For a one-off invocation without installing the command globally:

uvx zenthrix --version

System Requirements

Platform	Minimum Version
macOS	14.0+ (Apple Silicon M1/M2/M3/M4)
Linux	Ubuntu 22.04+ (aarch64 / x86_64)
Android	NDK r25+ (for targeting Snapdragon platforms)

Quickstart

1. Compile a Model

Compilation requires the separately distributed native engine. Without it, the CLI reports an actionable error rather than producing an invalid .zx file.

Compile an ONNX or GGUF model targeting local hardware execution:

zenthrix compile \
  --model meta-llama/Llama-3.2-1B-Instruct \
  --format onnx \
  --target auto \
  --quantization int4 \
  --output ./llama-3.2-1b.zx

2. Inspect Graph Optimizations

Analyze operator fusions and projected memory footprints prior to compilation:

zenthrix inspect ./llama-3.2-1b.zx --memory-profile

3. Run Inference via CLI

Verify compiled throughput directly in your terminal:

zenthrix run \
  --model ./llama-3.2-1b.zx \
  --prompt "Explain quantum decoherence in two sentences." \
  --max-tokens 128

4. Python API Usage

import zenthrix

# Load and initialize the compiled runtime
engine = zenthrix.Engine(model_path="./llama-3.2-1b.zx")

# Execute a deterministic inference pass
output = engine.generate(
    prompt="Synthesize the primary risks of high inference latency.",
    temperature=0.2,
    max_tokens=256,
)

print(output.text)
print(f"Time to First Token (TTFT): {output.ttft_ms} ms")
print(f"Throughput: {output.tokens_per_second} tokens/sec")

Automation and Runtime Availability

The public package does not include the proprietary native compiler/runtime. compile, inspect, and run therefore return a non-zero status until a compatible runtime is provisioned; they never create placeholder artifacts or claim successful inference.

For automation, add --json to run, inspect, or compile. Expected failures are emitted as a JSON object with error and message fields:

{"error": "EngineUnavailableError", "message": "..."}

Successful inference uses text, ttft_ms, and tokens_per_second fields.

Supported Target Architectures

Silicon Target	Optimization Backend	Compute Units
Apple Silicon (M-Series / A-Series)	Metal MSL & AMX Matrix Intrinsics	GPU / Neural Engine
Qualcomm Snapdragon (8 Gen 2/3/4)	Hexagon HTP Architecture (C++)	NPU / HVX
Arm Neoverse / Cortex	Arm NEON / SVE2 Assembly	CPU Vector Extensions

Contributing

We welcome community contributions to adapters, loaders, and frontend parsers. All contributions require signing our Contributor License Agreement (CLA) during the pull request process.

See CONTRIBUTING.md for local environment setup instructions.

License

The Zenthrix CLI and client adapters are distributed under the Apache License 2.0. The underlying compilation engine dynamic binary is subject to the WithBrian Technologies Commercial EULA embedded in binary distributions.