> For the complete documentation index, see [llms.txt](https://docs.mind-mesh.info/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.mind-mesh.info/mindmesh/architecture.md).

# Architecture

Mindmesh is composed of five primary modules that work in tandem to form a fully decentralized AI service layer:

## 4.1 Compute Mesh

A distributed GPU network where node operators provide compute capacity by staking $MESH and accepting tasks submitted by users or model developers. Each node runs containerized workloads, metered by smart contracts. Nodes are rewarded based on uptime, performance, and trust score — enabling scalable and censorship-resistant compute infrastructure without relying on AWS or GCP.

## 4.2 Model Hub

A decentralized repository where developers upload, register, and version AI models (e.g. LLMs, vision models, agents). Each model is assigned a unique ID, can be token-gated, and is monetized per call or subscription. Model weights can be stored via IPFS or Filecoin, and inference rights are governed by on-chain permissions.

## 4.3 Data Grid

The Mindmesh Data Grid is an open marketplace for high-value training datasets. Contributors can tokenize datasets, define license rules, and receive $MESH rewards when data is used for training or inference. Future upgrades will support privacy-preserving data contribution and synthetic data generation.

## 4.4 Prompt Station

An interface layer where users query AI models using natural language. Prompts can be sent via API or UI, and responses are processed by models hosted in the Mesh. Prompt creators and AI agents can be monetized, and the interface supports version tracking, audit logs, and cost estimation.

## 4.5 Proof-of-Inference

To guarantee that off-chain model execution is trustworthy, Mindmesh implements a zkML-based system for **Proof-of-Inference**. This mechanism enables any model output to be verifiably traced to a specific model version, dataset, and input — without exposing private data or model internals. This is essential for high-stakes use cases like finance, healthcare, or governance.

Together, these components form a vertically integrated stack for AI services — owned by no one, accessible to all.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.mind-mesh.info/mindmesh/architecture.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.