# opendatalab/mineru-ecosystem architecture

> Architecture diagram of the GitHub repository opendatalab/mineru-ecosystem, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-02). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/opendatalab/mineru-ecosystem
- Repository: https://github.com/opendatalab/mineru-ecosystem

## Overview

MinerU-Ecosystem is a collection of clients and integrations for MinerU’s document-parsing API. Users submit files or URLs through a CLI, SDK, MCP tool, or framework loader; the service returns parsed Markdown or structured results, with precision and quick/Flash modes documented. The graph highlights the Python and TypeScript SDK internals, the MCP request path, and the LangChain and LlamaIndex integrations. Some CLI and integration internals were not excerpted, so only documented or indexed relationships are shown.

Main entry points: [`sdk/python/src/mineru/client.py`](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/client.py) (Python client), [`sdk/typescript/src/client.ts`](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/client.ts) (TypeScript client), [`mcp/src/mineru_open_mcp/tools/extract.py`](https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/tools/extract.py) (MCP extraction), [`cli/mineru-open-api/cmd`](https://github.com/opendatalab/mineru-ecosystem/tree/main/cli/mineru-open-api/cmd) (CLI commands)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_cli["Command line"]
  node_cli["CLI commands"]
  node_cli_entry["CLI entrypoint<br/>[main.go]"]
end

subgraph group_sdks["Client SDKs"]
  node_go_sdk["Go SDK<br/>[client.go]"]
  node_py_client["Python client<br/>[client.py]"]
  node_py_api["Precision API<br/>[_api.py]"]
  node_py_flash["Flash API<br/>[_flash_api.py]"]
  node_py_models["Python results<br/>[models.py]"]
  node_py_zip["ZIP parsing<br/>[_zip.py]"]
  node_py_errors["Python errors<br/>[exceptions.py]"]
  node_ts_client["TypeScript client<br/>[client.ts]"]
  node_ts_api["Precision API<br/>[api.ts]"]
  node_ts_flash["Flash API<br/>[flash-api.ts]"]
  node_ts_models["TypeScript results<br/>[models.ts]"]
  node_ts_zip["ZIP parsing<br/>[zip.ts]"]
  node_ts_errors["TypeScript errors<br/>[errors.ts]"]
end

subgraph group_agent["Agent access"]
  node_mcp_server["MCP server<br/>[server.py]"]
  node_mcp_tools["MCP tools<br/>[tools.py]"]
  node_mcp_extract["MCP extraction<br/>[extract.py]"]
  node_mcp_config["MCP configuration<br/>[config.py]"]
end

subgraph group_rag["RAG integrations"]
  node_langchain_loader["LangChain loader<br/>[mineru.py]"]
  node_pdf_tools["PDF utilities<br/>[pdf.py]"]
  node_llama_reader["LlamaIndex reader<br/>[base.py]"]
end

subgraph group_service["Parsing service"]
  node_api{{"MinerU Open API"}}
  node_markdown["Markdown and results"]
end

node_user(("Developer or agent"))

node_user -->|"runs"| node_cli_entry
node_cli_entry -->|"executes"| node_cli
node_cli -->|"configures client"| node_go_sdk
node_cli -->|"submits parse and crawl"| node_api
node_user -->|"calls"| node_py_client
node_py_client -->|"uses"| node_py_api
node_py_client -->|"uses"| node_py_flash
node_py_client -->|"builds results"| node_py_models
node_py_client -->|"parses downloads"| node_py_zip
node_py_client -->|"raises"| node_py_errors
node_py_api -->|"maps errors"| node_py_errors
node_py_client -->|"submits and polls"| node_api
node_api -->|"returns parsed content"| node_markdown
node_user -->|"calls"| node_ts_client
node_ts_client -->|"uses"| node_ts_api
node_ts_client -->|"uses"| node_ts_flash
node_ts_client -->|"builds results"| node_ts_models
node_ts_client -->|"parses downloads"| node_ts_zip
node_ts_client -->|"raises"| node_ts_errors
node_ts_api -->|"maps errors"| node_ts_errors
node_ts_client -->|"submits and polls"| node_api
node_user -->|"connects"| node_mcp_server
node_mcp_server -->|"registers tools"| node_mcp_tools
node_mcp_tools -->|"dispatches extraction"| node_mcp_extract
node_mcp_extract -->|"calls SDK"| node_py_client
node_mcp_server -->|"reads settings"| node_mcp_config
node_mcp_extract -->|"uses configuration"| node_mcp_config
node_mcp_extract -->|"saves parsed output"| node_markdown
node_user -->|"loads documents"| node_langchain_loader
node_langchain_loader -->|"uses PDF helpers"| node_pdf_tools
node_langchain_loader -->|"parses sources"| node_api
node_user -->|"loads documents"| node_llama_reader
node_llama_reader -->|"parses sources"| node_api
node_go_sdk -->|"submits parse requests"| node_api

click node_cli "https://github.com/opendatalab/mineru-ecosystem/tree/main/cli/mineru-open-api/cmd"
click node_cli_entry "https://github.com/opendatalab/mineru-ecosystem/blob/main/cli/mineru-open-api/main.go"
click node_go_sdk "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/go/client.go"
click node_py_client "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/client.py"
click node_py_api "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/_api.py"
click node_py_flash "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/_flash_api.py"
click node_py_models "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/models.py"
click node_py_zip "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/_zip.py"
click node_py_errors "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/exceptions.py"
click node_ts_client "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/client.ts"
click node_ts_api "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/api.ts"
click node_ts_flash "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/flash-api.ts"
click node_ts_models "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/models.ts"
click node_ts_zip "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/zip.ts"
click node_ts_errors "https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/errors.ts"
click node_mcp_server "https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/server.py"
click node_mcp_tools "https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/tools/tools.py"
click node_mcp_extract "https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/tools/extract.py"
click node_mcp_config "https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/config.py"
click node_langchain_loader "https://github.com/opendatalab/mineru-ecosystem/blob/main/langchain_mineru/langchain_mineru/document_loaders/mineru.py"
click node_pdf_tools "https://github.com/opendatalab/mineru-ecosystem/blob/main/langchain_mineru/langchain_mineru/utils/pdf.py"
click node_llama_reader "https://github.com/opendatalab/mineru-ecosystem/blob/main/llama-index-readers-mineru/llama_index/readers/mineru/base.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_cli,node_cli_entry toneBlue
class node_go_sdk,node_py_client,node_py_api,node_py_flash,node_py_models,node_py_zip,node_py_errors,node_ts_client,node_ts_api,node_ts_flash,node_ts_models,node_ts_zip,node_ts_errors toneAmber
class node_mcp_server,node_mcp_tools,node_mcp_extract,node_mcp_config toneMint
class node_langchain_loader,node_pdf_tools,node_llama_reader toneRose
class node_api,node_markdown,node_user toneIndigo
```

## Components

### Command line

- [CLI commands](https://github.com/opendatalab/mineru-ecosystem/tree/main/cli/mineru-open-api/cmd): `cli/mineru-open-api/cmd`
- [CLI entrypoint](https://github.com/opendatalab/mineru-ecosystem/blob/main/cli/mineru-open-api/main.go): `cli/mineru-open-api/main.go`

### Client SDKs

- [Go SDK](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/go/client.go): `sdk/go/client.go`
- [Python client](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/client.py): `sdk/python/src/mineru/client.py`
- [Precision API](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/_api.py): `sdk/python/src/mineru/_api.py`
- [Flash API](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/_flash_api.py): `sdk/python/src/mineru/_flash_api.py`
- [Python results](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/models.py): `sdk/python/src/mineru/models.py`
- [ZIP parsing](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/_zip.py): `sdk/python/src/mineru/_zip.py`
- [Python errors](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/python/src/mineru/exceptions.py): `sdk/python/src/mineru/exceptions.py`
- [TypeScript client](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/client.ts): `sdk/typescript/src/client.ts`
- [Precision API](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/api.ts): `sdk/typescript/src/api.ts`
- [Flash API](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/flash-api.ts): `sdk/typescript/src/flash-api.ts`
- [TypeScript results](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/models.ts): `sdk/typescript/src/models.ts`
- [ZIP parsing](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/zip.ts): `sdk/typescript/src/zip.ts`
- [TypeScript errors](https://github.com/opendatalab/mineru-ecosystem/blob/main/sdk/typescript/src/errors.ts): `sdk/typescript/src/errors.ts`

### Agent access

- [MCP server](https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/server.py): `mcp/src/mineru_open_mcp/server.py`
- [MCP tools](https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/tools/tools.py): `mcp/src/mineru_open_mcp/tools/tools.py`
- [MCP extraction](https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/tools/extract.py): `mcp/src/mineru_open_mcp/tools/extract.py`
- [MCP configuration](https://github.com/opendatalab/mineru-ecosystem/blob/main/mcp/src/mineru_open_mcp/config.py): `mcp/src/mineru_open_mcp/config.py`

### RAG integrations

- [LangChain loader](https://github.com/opendatalab/mineru-ecosystem/blob/main/langchain_mineru/langchain_mineru/document_loaders/mineru.py): `langchain_mineru/langchain_mineru/document_loaders/mineru.py`
- [PDF utilities](https://github.com/opendatalab/mineru-ecosystem/blob/main/langchain_mineru/langchain_mineru/utils/pdf.py): `langchain_mineru/langchain_mineru/utils/pdf.py`
- [LlamaIndex reader](https://github.com/opendatalab/mineru-ecosystem/blob/main/llama-index-readers-mineru/llama_index/readers/mineru/base.py): `llama-index-readers-mineru/llama_index/readers/mineru/base.py`

### Parsing service

- **MinerU Open API**
- **Markdown and results**

### Other components

- **Developer or agent**

## Connections

- Developer or agent → CLI entrypoint: runs (evidence: `README.md`)
- CLI entrypoint → CLI commands: executes (evidence: `cli/mineru-open-api/main.go`)
- CLI commands → Go SDK: configures client (evidence: `cli/mineru-open-api/cmd/client.go`)
- CLI commands → MinerU Open API: submits parse and crawl (evidence: `README.md`)
- Developer or agent → Python client: calls (evidence: `README.md`)
- Python client → Precision API: uses (evidence: `sdk/python/src/mineru/client.py`)
- Python client → Flash API: uses (evidence: `sdk/python/src/mineru/client.py`)
- Python client → Python results: builds results (evidence: `sdk/python/src/mineru/client.py`)
- Python client → ZIP parsing: parses downloads (evidence: `sdk/python/src/mineru/client.py`)
- Python client → Python errors: raises (evidence: `sdk/python/src/mineru/client.py`)
- Precision API → Python errors: maps errors (evidence: `sdk/python/src/mineru/_api.py`)
- Python client → MinerU Open API: submits and polls (evidence: `README.md`)
- MinerU Open API → Markdown and results: returns parsed content (evidence: `README.md`)
- Developer or agent → TypeScript client: calls (evidence: `README.md`)
- TypeScript client → Precision API: uses (evidence: `sdk/typescript/src/client.ts`)
- TypeScript client → Flash API: uses (evidence: `sdk/typescript/src/client.ts`)
- TypeScript client → TypeScript results: builds results (evidence: `sdk/typescript/src/client.ts`)
- TypeScript client → ZIP parsing: parses downloads (evidence: `sdk/typescript/src/client.ts`)
- TypeScript client → TypeScript errors: raises (evidence: `sdk/typescript/src/client.ts`)
- Precision API → TypeScript errors: maps errors (evidence: `sdk/typescript/src/api.ts`)
- TypeScript client → MinerU Open API: submits and polls (evidence: `README.md`)
- Developer or agent → MCP server: connects (evidence: `README.md`)
- MCP server → MCP tools: registers tools (evidence: `mcp/src/mineru_open_mcp/server.py`)
- MCP tools → MCP extraction: dispatches extraction (evidence: `mcp/src/mineru_open_mcp/tools/tools.py`)
- MCP extraction → Python client: calls SDK (evidence: `mcp/src/mineru_open_mcp/tools/extract.py`)
- MCP server → MCP configuration: reads settings (evidence: `mcp/src/mineru_open_mcp/server.py`)
- MCP extraction → MCP configuration: uses configuration (evidence: `mcp/src/mineru_open_mcp/tools/extract.py`)
- MCP extraction → Markdown and results: saves parsed output (evidence: `mcp/src/mineru_open_mcp/tools/extract.py`)
- Developer or agent → LangChain loader: loads documents (evidence: `README.md`)
- LangChain loader → PDF utilities: uses PDF helpers (evidence: `langchain_mineru/langchain_mineru/document_loaders/mineru.py`)
- LangChain loader → MinerU Open API: parses sources (evidence: `README.md`)
- Developer or agent → LlamaIndex reader: loads documents (evidence: `README.md`)
- LlamaIndex reader → MinerU Open API: parses sources (evidence: `README.md`)
- Go SDK → MinerU Open API: submits parse requests (evidence: `README.md`)
