NVIDIA NeMo Fabric

NeMo Fabric gives users one configurable, observable way to run applications across multiple agent harnesses. It standardizes configuration, lifecycle management, and results without requiring a separate integration for every harness.

NeMo Fabric lets you change harnesses without rebuilding each integration, isolate conflicting runtime dependencies, and manage harness configuration, execution, and observability consistently. Every run returns normalized results, artifacts, and telemetry for downstream systems to consume.

It provides:

• a versioned, typed configuration contract;
• ordinary Python composition for experiment variants;
• adapter integrations for harness-specific launch and control;
• a Python SDK backed by the Rust core;
• normalized run results, artifact manifests, and telemetry references.

Install

Install the core runtime and Python SDK:

pip install nemo-fabric

Supported Python Versions

NeMo Fabric supports Python 3.11 through 3.14. However, some harnesses and integrations have more restrictive requirements. Hermes Agent requires Python 3.11 through 3.13, and the Harbor integration requires Python 3.12 or later.

Supported Harnesses

The following table shows the install target for each supported agent harness:

To install the NeMo Fabric runtime, adapter, and supported harness in one environment, choose one of the following nemo-fabric harness extras:

pip install "nemo-fabric[claude]"
pip install "nemo-fabric[codex]"
pip install "nemo-fabric[deepagents]"
pip install "nemo-fabric[mini-swe-agent]"

Hermes Agent 0.20 and later is not installable from PyPI. For this reason the Hermes Agent adapter does not provide a harness extra.Follow the Hermes Agent installation guide, then install nemo-fabric and nemo-fabric-adapters-hermes into the Python environment that runs Hermes Agent.

To install an adapter and its harness without the NeMo Fabric runtime, choose one of the following adapter package harness extras:

pip install "nemo-fabric-adapters-claude[harness]"
pip install "nemo-fabric-adapters-codex[harness]"
pip install "nemo-fabric-adapters-deepagents[harness]"
pip install "nemo-fabric-adapters-mini-swe-agent[harness]"

Every adapter package also provides an adapter-scoped full extra, which does not install the NeMo Fabric runtime. For Claude, Codex, and mini-SWE-agent, full installs the same dependencies as harness. For LangChain Deep Agents, full also installs the NeMo Relay Python package. The Hermes adapter's full extra installs NeMo Relay but does not install Hermes Agent.

If the environment already manages a compatible harness, choose one of the following bare adapter packages:

pip install nemo-fabric-adapters-claude
pip install nemo-fabric-adapters-codex
pip install nemo-fabric-adapters-deepagents
pip install nemo-fabric-adapters-hermes
pip install nemo-fabric-adapters-mini-swe-agent

The adapter distribution contains only adapter-owned runtime dependencies. It does not install the NeMo Fabric runtime. For package-installable harnesses, select harness or full to install the harness. The Hermes adapter never installs Hermes Agent. If the runtime shares an environment with an existing compatible harness, install nemo-fabric and the bare adapter package together.

Integrations

Harbor Integration

Install the Harbor integration with the harbor extra:

pip install "nemo-fabric[harbor]"

NeMo Relay Integration

Install a compatible NeMo Relay package with the relay extra:

pip install "nemo-fabric[relay]"

This installs a version of the NeMo Relay Python package known to be compatible with the installed version of NeMo Fabric.

The LangChain Deep Agents and Hermes Agent adapter packages also provide relay and full extras for environments that do not install nemo-fabric. Claude and Codex require the nemo-relay CLI instead of the NeMo Relay Python package. They do not provide a relay extra. Refer to the NeMo Relay CLI install guide for instructions on installing the CLI tool.

Learn More

Refer to the NVIDIA NeMo Fabric documentation for installation, configuration, and usage guidance. Source code is available in the NVIDIA NeMo Fabric repository.