Why is Neural Inverse open source?
We believe open source is the right foundation for developer tooling, especially for software that runs inside regulated and safety-critical environments:
- Transparency and Trust: Every line of code that touches your firmware, your legacy codebase, or your compliance data can be inspected. No black boxes. No hidden telemetry.
- No vendor lock-in: Neural Inverse runs entirely on your machine. Your LLM API keys never leave your environment. Switch models, switch providers, or run fully offline at any time.
- Community-driven development: Integrations, hardware support, and language pairs grow faster when the community can contribute directly. The 357+ MCU variants and 36 migration language pairs we support today came from exactly this model.
- Sovereignty: Teams in air-gapped environments, regulated industries, and on-premise deployments can run Neural Inverse without any network dependency after the initial install.
- Extensibility: The open source edition is the full product. Nothing is artificially limited to push you toward a paid tier.
Licensing Overview
The Neural Inverse open source edition is licensed under Apache 2.0 — an OSI-approved permissive license. You can use it freely in commercial products, modify it, fork it, and distribute it.
The full open source edition includes Power Mode agentic coding (Explorer, Editor, Verifier sub-agents), firmware development (357+ MCU variants, SVD register auto-load), legacy modernization (COBOL, FORTRAN, RPG, and 36 migration language pairs), and BYOLLM support. No usage limits. No telemetry. No credit card required.
Enterprise add-ons (Checks Engine GRC enforcement, IAM Engine, Enclave air-gapped deployment, Management Console, cryptographic audit trail) require a commercial Enterprise license.
Do you have any questions? Reach out via GitHub Discussions.
Open Source vs Enterprise
| Capability | Open Source (Apache 2.0) | Enterprise |
|---|---|---|
| Power Mode (Explorer, Editor, Verifier) | Yes | Yes |
| Firmware dev (357+ MCU variants, SVD) | Yes | Yes |
| Legacy modernization (36 language pairs) | Yes | Yes |
| BYOLLM (Claude, GPT-4, Gemini, Ollama, Bedrock) | Yes | Yes |
| Checks Engine (GRC / compliance enforcement) | - | Yes |
| IAM Engine (SSO, RBAC, audit) | - | Yes |
| Enclave (air-gapped / on-prem deployment) | - | Yes |
| Management Console (fleet-wide policy) | - | Yes |
| Cryptographic audit trail | - | Yes |
| Support SLO and onboarding | - | Yes |
Open Source Edition
The open source edition is the complete Neural Inverse IDE. Install it with a single command:
curl -fsSL https://cdn.neuralinverse.io/ce/install.sh | bashOr on Windows:
irm https://neuralinverse.com/win | iexAll product capabilities are available immediately. No account, no API key, no credit card.
Enterprise Edition
Enterprise add-ons are shipped in clearly marked directories within the same repository. They are visible as source code but require a commercial license key to activate. The core open source edition works 100% without them — they are purely additive.
Enterprise is designed for teams that need to enforce compliance frameworks (SOC 2, MISRA C, IEC 62304, ISO 26262, DO-178C, PCI-DSS, IEC 62443) at the IDE level, with auditable proof of enforcement.
See Enterprise or contact sales@neuralinverse.com.
Frequently Asked Questions
Is Neural Inverse really open source?
Yes. The full open source edition is Apache 2.0 licensed — an OSI-approved license. You get the four freedoms: use, study, modify, distribute. There is no "open core" that locks features behind a commercial gate within the open source edition itself.
What is BYOLLM and does it mean my keys stay local?
BYOLLM (Bring Your Own LLM) means you connect Neural Inverse to any LLM provider using your own API keys. Those keys are stored locally on your machine and are never transmitted to Neural Inverse servers. Supported providers include Anthropic Claude, OpenAI GPT-4, Google Gemini, Ollama (local models), and AWS Bedrock.
Can I run Neural Inverse completely offline?
Yes. After installation, Neural Inverse has no outbound network dependency. Teams in air-gapped and classified environments run it on isolated networks with local LLM endpoints (via Ollama or similar). No telemetry is sent by default. The open source edition has no phone-home behavior.
What does the Enterprise Enclave option provide?
Enclave is an air-gapped, on-prem deployment of Neural Inverse Enterprise with no cloud dependency whatsoever. It includes the full Enterprise add-on suite (Checks Engine, IAM Engine, Management Console, cryptographic audit trail) deployed entirely within your own infrastructure. It is designed for defense, aerospace, and regulated industries where data must never leave a controlled environment.
Which self-host licensing options exist?
The open source edition is free forever under Apache 2.0 and requires no license key. Enterprise add-ons require a commercial license key. Contact sales@neuralinverse.com for Enterprise pricing.
What compliance frameworks does the Checks Engine cover?
The Checks Engine enforces rules for: SOC 2, MISRA C, IEC 62304 (medical device software), ISO 26262 (automotive), DO-178C (aviation), PCI-DSS (payment card), and IEC 62443 (industrial control systems). Rule sets are configurable per project. Evidence is stored in the cryptographic audit trail.
How do I contribute?
The best ways to contribute:
- Submit and vote on ideas via GitHub Discussions.
- File or comment on Issues.
- Open a pull request. See
CONTRIBUTING.mdin the repository to get started.
I want to see the code.
The repository is at github.com/NeuralInverse/neuralinverse.
