MANGOS
MANGOSMANGOS Explained: A Systems Map of Meta, Anthropic, Nvidia, Google, OpenAI and SpaceX
An original framework for tracking six influential AI-era companies by role in the stack—models, distribution, compute, platforms, connectivity, capital intensity, and execution risk.
The MANGOS systems map
The roles overlap. Google and Meta own infrastructure. Nvidia has software and systems. Anthropic and OpenAI increasingly ship agent platforms. SpaceX is not a frontier-model company in the same way, but connectivity and large-scale infrastructure make it useful to the broader systems lens.
Seven dimensions to compare
| Dimension | Question | Evidence |
|---|---|---|
| Model capability | How strong and useful are current models? | Official model docs and eval methodology |
| Distribution | How easily can users adopt the product? | Installed base, platforms, channels |
| Developer ecosystem | APIs, SDKs, tools, integrations | Documentation and usage evidence |
| Compute position | Owned or supplied infrastructure | Capex, hardware, cloud footprint |
| Enterprise position | Identity, controls, support, procurement | Enterprise product docs |
| Capital intensity | How much infrastructure is required? | Filings and capex |
| Execution risk | Regulation, supply, product, concentration | Filings, regulators, primary reporting |
Read each company through a different lens
Meta
Watch the interaction between consumer distribution, open model strategy, advertising economics, and infrastructure investment.
Anthropic
Watch model quality, Claude Code, enterprise controls, tool use, MCP, and the company’s ability to convert technical strength into durable enterprise workflows.
Nvidia
Watch accelerator demand, networking, systems, software ecosystem, supply chain, customer concentration, and whether installed capacity remains well utilized.
Watch the combination of Gemini, Cloud, search/distribution, productivity software, custom accelerators, and developer platforms.
OpenAI
Watch application distribution, developer platform, enterprise adoption, agent workflows, compute economics, and the relationship between product expansion and capital needs.
SpaceX
Watch connectivity, launch economics, infrastructure scale, enterprise/government exposure, and how public-company reporting changes the evidence available to outside analysts.
Separate signal from narrative
For each company, write the signal in one sentence, cite the primary evidence, state your confidence, and describe what would invalidate the conclusion. For example, strong quarterly infrastructure revenue is evidence of current demand, but it does not prove that growth will persist indefinitely.
SpaceX is now publicly reportable through SEC filings, which means XTIANZ can use filings rather than treating the company as a private-market proxy. This is exactly why the framework needs review dates: the evidence landscape changes.
How to use the framework
Use MANGOS to ask cross-company questions. If model capability converges, who has distribution? If compute becomes constrained, who controls supply? If inference becomes cheaper, which businesses benefit from more usage? If enterprise governance becomes the bottleneck, which platforms can meet identity, audit, and procurement requirements?
The framework is most useful when it creates falsifiable questions rather than a list of favorite companies.
Review history
August 9, 2026 — Reworked as a flagship XTIANZ guide with current primary sources, original decision frameworks, and technical review.
Suggest a correction ↗Disclosure
AI tools may assist research organization, drafting, code, and quality checks. The final structure, claims, frameworks, and publication decision are manually reviewed. XTIANZ does not accept payment to change technical conclusions.