Flexible deployment from personal laptops to air-gapped research clusters.
Choose the edition that matches your deployment model, team size, compliance needs, and research intensity.
Available entry points
Solo
Lab
Enterprise
Plan comparison
Pricing & deployment summary
| Feature / Capability | Edition Individual Edition Free Forever | Edition Lab Suite: Trial Free (Evaluation) | Edition Lab Suite: Academic Subsidized / Grant-Friendly | Edition Lab Suite: Enterprise & Govt. Custom Commercial |
|---|---|---|---|---|
| Target Audience | Solo scientists, PhD students, independent researchers | Labs evaluating multi-agent workflows | Universities, academic institutes, non-profit labs | Defense, Pharma, Semis, Energy, Govt. R&D |
| Target Architecture | PCs & Laptops | Workstations / Distributed | Workstations & Clusters | Enterprise Server Stacks |
| Deployment Mode | Standalone local app (Single device) | Docker-based container images | Docker / Distributed multi-system | Distributed multi-node / High-density server stacks |
| Agent Architecture | Single-agent runtime | Multi-agent runtime (Basic orchestration) | Agentic Swarms | Agentic Swarms (Parallel high-throughput) |
| Admin & Access Control | Local user | Single admin control | Single admin + project roles | Single Admin + Multi-supervisor controls |
| Database & Knowledge Base | Single local vector store & DB | Shared database + discrete KB | Discrete KB + project-level DB | Discrete DB & KB per project & experiment |
| Supported Models | Open-weight GGUF (Granite, Gemma, Qwen) | Open-weight GGUF & on-prem open models | Open weights + research checkpoints | Proprietary on-prem models & custom inference plugins |
| Simulations & Modeling | Python execution (NumPy, SciPy) | Standard AI layer & code execution | PINNs + GNU Octave + C / C++ / R | Full Suite + World Model Simulations |
| Primary Action |
Detailed breakdown
Edition-by-edition details
1. Individual Edition — Free Forever
Designed for independent researchers, engineers, and graduate students running on their personal machines.
- Hardware (Minimum): 16GB RAM for CPU-only, or 8GB RAM + 4GB VRAM for small GPUs. Compatible with modern PCs, laptops, and Apple Silicon Macs.
- Core Stack: Standalone binary with a single-agent orchestrator.
- Execution: Local Python sandbox (NumPy/SciPy), native LaTeX compilation, and ChromaDB vector search.
- Model Support: Fully offline open weights (IBM Granite, Google Gemma, Qwen) running via quantized GGUF format.
- Privacy: Zero cloud connection, zero telemetry, 100% private.
2. Lab Suite — Free Trial Edition
Designed for research teams, PIs, and lab managers testing multi-system orchestration before institutional procurement.
- Hardware: Local lab workstations or distributed LAN setups.
- Deployment: Docker-based containerized environment linking multiple systems to a single database.
- Core AI Layer: Multi-agent reasoning loops, team-shared Knowledge Base Management System (KBMS), and central project management.
- Trial Scope: Unlimited access to the core AI layer and multi-agent coordination. Excludes World Model simulations and proprietary on-premise model inference plugins.
3. Lab Suite — Academic Edition
Discounted, grant-aligned annual licenses for universities, public research labs, and academic consortia.
- Everything in the Trial Edition, plus:
- Mathematical & Simulation Stack: Native integration with GNU Octave, plus execution and orchestration support for C, C++, and R.
- Physics-Informed Neural Networks (PINNs): Run physics-aware neural surrogate models for accelerated, lower-cost scientific simulations.
- Agentic Swarms: Orchestrate complex, multi-agent collaborative swarms running parallel reasoning tasks across lab workstations.
- Discrete KBMS: Isolated knowledge bases for separate grant projects, teams, and lab members.
4. Lab Suite — Enterprise & Government Edition
Tailored for defense, national laboratories, pharmaceutical R&D, and semiconductor teams with stringent IP and regulatory mandates.
- Everything in Academic, plus:
- Server Stack Architecture: High-density deployment across on-prem server clusters with parallel multi-agent operations.
- World Model Simulations: Long-horizon, deterministic world model-based simulation and modeling engines for high-precision physical/systems research.
- Proprietary Model Plugins: Specialized on-premise deployment adapters for internal proprietary foundation models and custom inference engines.
- Total Isolation Architecture: Discrete, encrypted database and vector knowledge base generated dynamically per project and per experiment.
- Governance & Access: Strict Single-Admin with Multi-Supervisor Hierarchical Controls to manage authorization, experiment review, and compliance auditing.
- Air-Gap Certification: 100% network-isolated deployment packages with on-prem license verification keys.
What powers the Lab Suite?
- Physics-Aware Neural Networks (PINNs): Replace heavy finite-element passes with physics-constrained neural models for fast, low-compute parameter sweeps.
- World Model Simulations: Build internal predictive dynamics of complex systems for robust, long-duration modeling.
- Multi-Language R&D Engine: Native orchestration pipelines for Python, C, C++, R, and GNU Octave in a unified execution canvas.
- Knowledge Base Management System (KBMS): Purpose-built scientific RAG architecture that indexes datasets, formulas, citations, and intermediate experiment logs.
- Agentic Swarms: Autonomous swarms built on LangGraph that divide literature search, script execution, hypothesis validation, and paper synthesis across specialized agents.
Choosing the right edition
Find the appropriate scale for your research workflows
Individual Researchers
The Individual Edition gives you a fully local, privacy-first workspace with no cloud dependency, ideal for laptops and workstations.
Academic & Lab Teams
The Trial and Academic editions unlock shared knowledge bases, distributed LAN orchestration, and scalable agentic swarms.
Enterprise & Government
Delivers air-gapped isolation, custom model plugins, world model simulations, and high-throughput cluster deployment.