### Minimal Setup for Running Paper2Any

Source: https://github.com/opendcai/paper2any/blob/main/docs/index.md

This snippet shows the minimal commands required to set up and start the Paper2Any application. It involves copying example environment files and then executing a start script.

```bash
cp fastapi_app/.env.example fastapi_app/.env
cp frontend-workflow/.env.example frontend-workflow/.env

bash deploy/start_nv.sh
```

--------------------------------

### Copy Example Configuration Files

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/configuration.md

Copy the example configuration files to their respective locations. The simple mode is recommended for initial setup.

```bash
cp fastapi_app/.env.simple.example fastapi_app/.env
cp frontend-workflow/.env.simple.example frontend-workflow/.env
cp deploy/profiles/nv.env.example deploy/profiles/nv.env
```

--------------------------------

### Deploy NVIDIA Machine

Source: https://github.com/opendcai/paper2any/blob/main/DEPLOY.md

Copy the example NVIDIA machine environment file and start the deployment script.

```bash
cp deploy/profiles/nv.env.example deploy/profiles/nv.env
bash deploy/start_nv.sh
```

--------------------------------

### Frontend Installation with npm

Source: https://github.com/opendcai/paper2any/blob/main/docs/installation.md

Navigates to the frontend directory and installs its dependencies using npm ci. Ensure Node.js and npm are installed and compatible.

```bash
cd frontend-workflow
npm ci
cd ..
```

--------------------------------

### Copy Backend Environment Example

Source: https://github.com/opendcai/paper2any/blob/main/DEPLOY.md

Copy the example backend environment file to create a new one for configuration.

```bash
cp fastapi_app/.env.example fastapi_app/.env
```

--------------------------------

### Start Backend Service Only

Source: https://github.com/opendcai/paper2any/blob/main/DEPLOY.md

Source the NVIDIA machine environment variables and then start the backend service.

```bash
set -a
source deploy/profiles/nv.env
set +a

bash deploy/start.sh
```

--------------------------------

### Backend Installation with Conda and Pip

Source: https://github.com/opendcai/paper2any/blob/main/docs/installation.md

Sets up the Python environment for the backend using conda, activates it, and installs project dependencies. Includes optional installation for NVIDIA GPU support and editable install.

```bash
conda create -n paper2any python=3.11 -y
conda activate paper2any

pip install --upgrade pip
pip install -r requirements-paper.txt
```

```bash
# NVIDIA GPU 机器再额外安装
pip install -r requirements-cu12.txt
```

```bash
pip install -e .
```

--------------------------------

### Copy Frontend Environment Example

Source: https://github.com/opendcai/paper2any/blob/main/DEPLOY.md

Copy the example frontend environment file to create a new one for configuration.

```bash
cp frontend-workflow/.env.example frontend-workflow/.env
```

--------------------------------

### Start Muxi Machine Deployment

Source: https://github.com/opendcai/paper2any/blob/main/DEPLOY.md

Execute the script to start the deployment on a Muxi machine.

```bash
bash deploy/start_muxi.sh
```

--------------------------------

### Linux Installation: Clone Repository and Install Base Dependencies

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Clones the Paper2Any repository and installs the base Python dependencies from requirements-base.txt. Also installs the project in editable mode.

```bash
# 1. Clone repository
git clone https://github.com/OpenDCAI/Paper2Any.git
cd Paper2Any

# 2. Install base dependencies
pip install -r requirements-base.txt

# 3. Install in editable (dev) mode
pip install -e .
```

--------------------------------

### Copy Example .env Files

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Copy the example environment files to create your own configuration files for the backend and frontend applications. These files are optional.

```bash
# Copy backend environment file
cp fastapi_app/.env.example fastapi_app/.env

# Copy frontend environment file
cp frontend-workflow/.env.example frontend-workflow/.env
```

--------------------------------

### Install Ubuntu/Debian System Packages

Source: https://github.com/opendcai/paper2any/blob/main/requirements-system-ubuntu.txt

Installs system packages listed in `requirements-system-ubuntu.txt` using `apt-get`. Ensure you run `apt-get update` first.

```bash
sudo apt-get update
sudo apt-get install -y $(grep -v '^#' requirements-system-ubuntu.txt | tr '\n' ' ')
```

--------------------------------

### Copy Deployment Profile Examples

Source: https://github.com/opendcai/paper2any/blob/main/deploy/profiles/README.md

Use these commands to copy example deployment profiles to local configuration files. Only fill in values your machine requires; empty values trigger auto-discovery.

```bash
cp deploy/profiles/muxi.env.example deploy/profiles/muxi.env
```

```bash
cp deploy/profiles/nv.env.example deploy/profiles/nv.env
```

--------------------------------

### Run Frontend with npm

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/open_source_deployment.md

Start the frontend development server using npm. Ensure you are in the correct directory and specify host and port.

```bash
cd frontend-workflow
npm run dev -- --host 0.0.0.0 --port 3000
```

--------------------------------

### Start Frontend Service Only

Source: https://github.com/opendcai/paper2any/blob/main/DEPLOY.md

Execute the script to start only the frontend service.

```bash
bash deploy/start_frontend.sh
```

--------------------------------

### Linux Installation: Install System Dependencies (Ubuntu)

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Installs essential system packages on Ubuntu, including ffmpeg, inkscape, libreoffice, poppler-utils, and wkhtmltopdf. These are external tools not installed by pip.

```bash
# 5. System dependencies (Ubuntu example; full list is mirrored in requirements-system-ubuntu.txt)
sudo apt-get update
sudo apt-get install -y ffmpeg inkscape libreoffice poppler-utils wkhtmltopdf
```

--------------------------------

### Start Local SAM3 Development

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/open_source_deployment.md

Use this script for single-machine development, which starts a SAM3 service, backend, and frontend. It's suitable for local debugging.

```bash
bash script/start_local_sam3_dev.sh
```

--------------------------------

### Create Conda Environment and Install Base Dependencies

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Sets up a new conda environment for Paper2Any and installs the core Python dependencies required for Windows.

```bash
# 0. Create and activate a conda environment
conda create -n paper2any python=3.12 -y
conda activate paper2any

# 1. Clone repository
git clone https://github.com/OpenDCAI/Paper2Any.git
cd Paper2Any

# 2. Install base dependencies
pip install -r requirements-win-base.txt

# 3. Install in editable (dev) mode
pip install -e .
```

--------------------------------

### Build and Run Docker Containers

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Builds and starts the backend and frontend Docker containers. Rebuild if code or .env files are changed.

```bash
# 3. Build + run
bash deploy/docker-up.sh
```

--------------------------------

### Run Backend with Uvicorn

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/open_source_deployment.md

Directly start the backend service using Uvicorn for development without deployment scripts. Ensure .env and dependencies are prepared.

```bash
uvicorn fastapi_app.main:app --host 0.0.0.0 --port 8000
```

--------------------------------

### Copy Muxi Machine Deployment Configuration

Source: https://github.com/opendcai/paper2any/blob/main/DEPLOY.md

Copy the example Muxi machine environment file to create a new one for machine-specific parameters.

```bash
cp deploy/profiles/muxi.env.example deploy/profiles/muxi.env
```

--------------------------------

### Start Local Model Service Cluster

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Use this script to start a local model service cluster for MinerU, SAM3, and OCR in a high-concurrency environment. Modify gpu_id and instance count as needed.

```bash
bash script/start_model_servers.sh
```

--------------------------------

### Launch Paper2Any Web Frontend

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Starts the Paper2Any web application using a one-click entrypoint script, recommended for NVIDIA machines. The frontend will be accessible at http://localhost:3000.

```bash
# Recommended one-click entrypoint on NVIDIA machines
bash deploy/start_nv.sh
```

--------------------------------

### Linux Installation: Resolve Dependency Conflicts

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Installs the doclayout_yolo package without its dependencies to avoid potential conflicts with other project requirements.

```bash
# 4. Resolve doclayout_yolo dependency conflicts (Important)
pip install doclayout_yolo --no-deps
```

--------------------------------

### Install vLLM for Local Inference Acceleration

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Installs a specific Windows build of vLLM for optional local inference acceleration. Ensure the wheel file matches your Python version, platform, and CUDA toolkit version.

```bash
pip install vllm-0.11.0+cu124-cp312-cp312-win_amd64.whl
```

--------------------------------

### Clone Paper2Any Repository and Configure Environment

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Clone the Paper2Any repository and set up the initial environment variables for Docker deployment. This includes copying example configuration files.

```bash
# 1. Clone
git clone https://github.com/OpenDCAI/Paper2Any.git
cd Paper2Any

# 2. Configure environment variables
cp fastapi_app/.env.simple.example fastapi_app/.env
cp frontend-workflow/.env.simple.example frontend-workflow/.env
cp deploy/docker.env.example deploy/docker.env
```

--------------------------------

### Install System Dependencies on Linux

Source: https://github.com/opendcai/paper2any/blob/main/docs/installation.md

Installs essential system packages required for various project functionalities, including graphics export and document conversion. Ensure you are in a Linux environment.

```bash
sudo apt-get update
sudo apt-get install -y \
  build-essential \
  curl \
  ffmpeg \
  git \
  inkscape \
  libreoffice \
  poppler-utils \
  wget \
  libgl1 \
  libglib2.0-0 \
  libgomp1 \
  libsm6 \
  libxext6 \
  libxrender1 \
  libsndfile1
```

--------------------------------

### Start Local SAM3 Service with Docker

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Starts the optional local SAM3 segmentation service using Docker Compose. This is required for PDF2PPT, Image2PPT, and Image2Drawio functionalities if not pointing to an external SAM3 service.

```bash
DOCKER_WITH_SAM3=1 bash deploy/docker-up.sh
```

--------------------------------

### Configure Environment Variables for Paper2Any

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Copy example environment files to active configuration files for simple or advanced modes. This is a necessary step before running the application.

```bash
cp fastapi_app/.env.simple.example fastapi_app/.env
cp frontend-workflow/.env.simple.example frontend-workflow/.env
```

```bash
cp fastapi_app/.env.example fastapi_app/.env
cp frontend-workflow/.env.example frontend-workflow/.env
```

--------------------------------

### Load MinerU Pre-trained Model with vLLM

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Starts a vLLM server to load the MinerU pre-trained model, specifying host, port, logits processor, and GPU memory utilization. This command is intended for PowerShell.

```powershell
# Start in PowerShell
vllm serve opendatalab/MinerU2.5-2509-1.2B \
  --host 127.0.0.1 \
  --port 8010 \
  --logits-processors mineru_vl_utils:MinerULogitsProcessor \
  --gpu-memory-utilization 0.6 \
  --trust-remote-code \
  --enforce-eager
```

--------------------------------

### Linux Installation: Install LaTeX Engine

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Installs the Tectonic LaTeX engine using Conda, which is recommended for LaTeX rendering.

```bash
# 3. LaTeX engine (tectonic) - recommended via conda
conda install -c conda-forge tectonic -y
```

--------------------------------

### Linux Installation: Install Paper-Specific Dependencies

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Installs Python dependencies required for paper processing, PDF/figure handling, and NVIDIA CUDA 12 extras (Linux only).

```bash
# 1. Paper / PDF / figure Python extras
pip install -r requirements-paper.txt

# 2. NVIDIA GPU runtime extras (Linux + CUDA 12 only)
pip install -r requirements-cu12.txt
```

--------------------------------

### Prepare Local Models

Source: https://github.com/opendcai/paper2any/blob/main/models/README.md

Run this script to populate the local models directory with necessary files. Ensure you have the script available in the 'script' directory.

```bash
bash script/prepare_local_models.sh
```

--------------------------------

### Install Paper2Any-Specific Dependencies

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Installs additional Python packages for LaTeX rendering and vector graphics processing, and configures the Tectonic LaTeX engine via conda.

```bash
# Python dependencies
pip install -r requirements-paper.txt

# tectonic: LaTeX engine (recommended via conda)
conda install -c conda-forge tectonic -y
```

--------------------------------

### Beautify PPT (Basic)

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Perform basic beautification on an existing PowerPoint file. Requires the input PPTX file, a style description, and an API key.

```bash
python script/run_ppt2polish_cli.py \
  --input old_presentation.pptx \
  --style "Academic style, clean and elegant" \
  --api-key sk-xxx
```

--------------------------------

### Frontend API Key and Base URL Configuration

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/configuration.md

Configure frontend settings such as the API key, base URL, and Supabase connection details.

```bash
VITE_API_KEY=
VITE_API_BASE_URL=
VITE_SUPABASE_URL=
VITE_SUPABASE_ANON_KEY=
```

--------------------------------

### Show Help Message

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Display detailed help and usage information for any of the CLI scripts. This is useful for understanding all available options and parameters.

```bash
python script/run_paper2figure_cli.py --help
```

--------------------------------

### Build Docker Images Only

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Builds the Docker images without starting the containers.

```bash
bash deploy/docker-build.sh
```

--------------------------------

### Backend API Key and Billing Mode Configuration

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/configuration.md

Configure essential backend settings including the API key, billing mode, and third-party API URLs.

```bash
BACKEND_API_KEY=
APP_BILLING_MODE=
DF_API_URL=
DF_API_KEY=
SUPABASE_URL=
SUPABASE_ANON_KEY=
SUPABASE_SERVICE_ROLE_KEY=
```

--------------------------------

### Docker Compose Configuration

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/open_source_deployment.md

docker-compose.yml file for orchestrating backend and frontend containers. Note that it does not automatically start SAM3/OCR clusters.

```yaml
docker-compose.yml
```

--------------------------------

### Configure Backend .env for Service Integration

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Configure external or local services like DrawIO OCR, MinerU, and SAM3 segmentation in the backend .env file. MinerU can fall back to a local port if the API key is empty.

```bash
# DrawIO OCR / VLM
PAPER2DRAWIO_OCR_API_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
PAPER2DRAWIO_OCR_API_KEY=your_dashscope_key

# MinerU official remote API; if MINERU_API_KEY is empty, backend falls back to local MINERU_PORT
MINERU_API_BASE_URL=https://mineru.net/api/v4
MINERU_API_KEY=your_mineru_api_key
MINERU_API_MODEL_VERSION=vlm

# SAM3 segmentation service for PDF2PPT / Image2PPT / Image2Drawio
# One endpoint:
SAM3_SERVER_URLS=http://127.0.0.1:8001
# Or multiple endpoints for load balancing:
# SAM3_SERVER_URLS=http://127.0.0.1:8021,http://127.0.0.1:8022
```

--------------------------------

### Beautify PPT (with Reference Image)

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Beautify an existing PowerPoint file using a reference image for style consistency. Requires the input PPTX, a style description, the reference image, and an API key.

```bash
python script/run_ppt2polish_cli.py \
  --input old_presentation.pptx \
  --style "Modern minimalist style" \
  --ref-img reference_style.png \
  --api-key sk-xxx
```

--------------------------------

### Run Paper2Drawio SAM3 Test

Source: https://github.com/opendcai/paper2any/blob/main/tests/_archive/manual/README.md

Execute the visual drawio workflow test for SAM3. This command is an example of how to run a specific manual test script.

```bash
python tests/_archive/manual/sam3/test_paper2drawio_sam3_back.py
```

--------------------------------

### Configure Frontend .env for Supabase

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Uncomment and configure Supabase URL and anon key in the frontend .env file if you want to enable user authentication features.

```bash
VITE_SUPABASE_URL=https://your-project.supabase.co
VITE_SUPABASE_ANON_KEY=your-anon-key
```

--------------------------------

### Prepare Local Models

Source: https://github.com/opendcai/paper2any/blob/main/docs/installation.md

A script to organize directories and download the RMBG-2.0 model. It does not automatically download the full SAM3 checkpoint or source code.

```bash
script/prepare_local_models.sh
```

--------------------------------

### Linux Installation: Create Conda Environment

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Creates and activates a new Conda environment named 'paper2any' with Python 3.11. This isolates project dependencies.

```bash
# 0. Create and activate a conda environment
conda create -n paper2any python=3.11 -y
conda activate paper2any
```

--------------------------------

### Deployment Profile Configuration

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/configuration.md

Configure deployment-specific parameters like Python paths, ports, and model server enablement.

```bash
APP_PYTHON=
PAPER2ANY_PYTHON=
APP_PORT=8000
FRONTEND_PORT=3000
SAM3_ENABLED=1
SAM3_GPUS=
```

--------------------------------

### Configure Backend .env for Supabase

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Configure Supabase credentials in the backend .env file if you require user authentication and cloud storage. This is optional.

```bash
SUPABASE_URL=https://your-project-id.supabase.co
SUPABASE_ANON_KEY=your_supabase_anon_key
```

--------------------------------

### Current Story Board JSON Example

Source: https://github.com/opendcai/paper2any/blob/main/dataflow_agent/toolkits/postertool/config/prompts/layout_balancer.txt

This JSON represents the current state of a poster's storyboard, detailing the arrangement of sections within each column, their content, and utilization.

```json
{
  "current_story_board": {
    "column_1": {
      "sections": ["intro_motivation", "related_work"],
      "utilization": "75%"
    },
    "column_2": {
      "sections": ["method_algorithm_a", "implementation_details"],
      "utilization": "90%"
    },
    "column_3": {
      "sections": ["results_summary", "conclusion"],
      "utilization": "98%"
    }
  }
}
```

--------------------------------

### Verify Runtime Configuration

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/open_source_deployment.md

Check the runtime configuration of the backend by sending a request to the /api/v1/account/runtime-config endpoint with an API key.

```bash
curl -H "X-API-Key: your-backend-api-key" \
  http://127.0.0.1:8000/api/v1/account/runtime-config
```

--------------------------------

### Column Analysis JSON Example

Source: https://github.com/opendcai/paper2any/blob/main/dataflow_agent/toolkits/postertool/config/prompts/layout_balancer.txt

This JSON provides an analysis of each column's current status, including its utilization percentage and a qualitative status (e.g., 'underutilized', 'optimal', 'overflow').

```json
{
  "column_analysis": {
    "left_utilization": "75%",
    "left_status": "underutilized",
    "middle_utilization": "90%",
    "middle_status": "optimal",
    "right_utilization": "98%",
    "right_status": "overflow",
    "available_height": "30 inches",
    "target_utilization": "85-95%"
  }
}
```

--------------------------------

### View Backend Logs

Source: https://github.com/opendcai/paper2any/blob/main/docs/guides/open_source_deployment.md

Execute a script to view the backend application logs.

```bash
bash deploy/logs.sh app
```

--------------------------------

### Structured Sections JSON Example

Source: https://github.com/opendcai/paper2any/blob/main/dataflow_agent/toolkits/postertool/config/prompts/layout_balancer.txt

This JSON structure represents available sections that can be added to a poster column. It includes details like section ID, title, content, importance level, and associated visual assets.

```json
{
  "structured_sections": [
    {
      "section_id": "intro_motivation",
      "section_title": "Motivation",
      "content": "The increasing complexity of modern systems necessitates novel approaches to resource management.",
      "importance_level": 2,
      "visual_assets": []
    },
    {
      "section_id": "method_algorithm_a",
      "section_title": "Algorithm A",
      "content": "Algorithm A employs a greedy approach to optimize task scheduling.",
      "importance_level": 1,
      "visual_assets": ["fig_algo_a.png"]
    }
  ]
}
```

--------------------------------

### Convert Paper to PPT Presentation (Basic)

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Basic usage for converting a PDF paper into a PowerPoint presentation. Requires an input PDF and an API key.

```bash
python script/run_paper2ppt_cli.py \
  --input paper.pdf \
  --api-key sk-xxx \
  --page-count 15
```

--------------------------------

### Convert Image to PPT (Basic)

Source: https://github.com/opendcai/paper2any/blob/main/README.md

Convert an image file into an editable PowerPoint presentation. Requires an input image file.

```bash
python script/run_image2ppt_cli.py --input screenshot.png
```