kreuzberg 3.3.0__tar.gz → 3.4.0__tar.gz

kreuzberg-3.4.0/.docker/Dockerfile

@@ -0,0 +1,21 @@
+FROM ghcr.io/astral-sh/uv:python3.13-bookworm as app
+ARG EXTRAS=""
+WORKDIR /app
+ENV PYTHONDONTWRITEBYTECODE 1
+ENV PYTHONUNBUFFERED 1
+ENV UV_LINK_MODE=copy
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    pandoc \
+    tesseract-ocr \
+    && apt-get clean && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
+
+
+COPY pyproject.toml uv.lock README.md ./
+COPY kreuzberg kreuzberg
+
+RUN uv sync --extra api${EXTRAS:+ --extra ${EXTRAS}} --no-editable --no-dev --compile-bytecode
+
+RUN groupadd -r appuser && useradd -r -g appuser -d /app -s /sbin/nologin appuser
+USER appuser
+CMD ["litestar", "--app", "kreuzberg._api.main:app", "run", "--host", "0.0.0.0"]

kreuzberg-3.4.0/.docker/README.md

@@ -0,0 +1,87 @@
+# Kreuzberg Docker Images
+
+[![GitHub](https://img.shields.io/badge/GitHub-Goldziher%2Fkreuzberg-blue)](https://github.com/Goldziher/kreuzberg)
+[![PyPI](https://badge.fury.io/py/kreuzberg.svg)](https://badge.fury.io/py/kreuzberg)
+[![Documentation](https://img.shields.io/badge/docs-GitHub_Pages-blue)](https://goldziher.github.io/kreuzberg/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Goldziher/kreuzberg/blob/main/LICENSE)
+
+High-performance Python library for text extraction from documents, available as Docker images.
+
+**Source Code**: [github.com/Goldziher/kreuzberg](https://github.com/Goldziher/kreuzberg)
+
+## Quick Start
+
+```bash
+docker run -p 8000:8000 goldziher/kreuzberg:latest
+```
+
+## Available Tags
+
+- `latest` - Latest stable release with API server and Tesseract OCR
+- `X.Y.Z` - Specific version (e.g., `3.0.0`)
+- `X.Y.Z-easyocr` - With EasyOCR support
+- `X.Y.Z-paddle` - With PaddleOCR support
+- `X.Y.Z-gmft` - With GMFT table extraction
+- `X.Y.Z-all` - With all optional dependencies
+
+## Usage
+
+### Extract Files via API
+
+```bash
+# Single file
+curl -X POST http://localhost:8000/extract \
+  -F "data=@document.pdf"
+
+# Multiple files
+curl -X POST http://localhost:8000/extract \
+  -F "data=@document1.pdf" \
+  -F "data=@document2.docx"
+```
+
+### Docker Compose
+
+```yaml
+version: '3.8'
+
+services:
+  kreuzberg:
+    image: goldziher/kreuzberg:latest
+    ports:
+      - "8000:8000"
+    restart: unless-stopped
+```
+
+## Features
+
+- **🚀 High Performance**: Optimized for speed and efficiency
+- **📄 Multiple Formats**: PDF, DOCX, images, HTML, and more
+- **🔍 OCR Support**: Built-in Tesseract, optional EasyOCR/PaddleOCR
+- **📊 Table Extraction**: Extract tables with GMFT
+- **🔒 Secure**: Runs as non-root user, no external API calls
+- **📦 Ready to Use**: Pre-configured API server
+
+## Documentation
+
+- **[GitHub Repository](https://github.com/Goldziher/kreuzberg)** - Source code and issue tracking
+- **[Full Documentation](https://goldziher.github.io/kreuzberg/)** - Complete user guide and API reference
+- **[API Documentation](https://goldziher.github.io/kreuzberg/user-guide/api-server/)** - REST API endpoints and usage
+- **[Docker Guide](https://goldziher.github.io/kreuzberg/user-guide/docker/)** - Detailed Docker usage guide
+
+## Support
+
+- **Issues**: [github.com/Goldziher/kreuzberg/issues](https://github.com/Goldziher/kreuzberg/issues)
+- **Discussions**: [github.com/Goldziher/kreuzberg/discussions](https://github.com/Goldziher/kreuzberg/discussions)
+- **Discord**: [Join our community](https://discord.gg/pXxagNK2zN)
+
+## Contributing
+
+Contributions are welcome! See our [Contributing Guide](https://github.com/Goldziher/kreuzberg/blob/main/docs/contributing.md).
+
+## License
+
+MIT License - see [LICENSE](https://github.com/Goldziher/kreuzberg/blob/main/LICENSE) for details.
+
+______________________________________________________________________
+
+Made with ❤️ by the [Kreuzberg contributors](https://github.com/Goldziher/kreuzberg/graphs/contributors)

kreuzberg-3.4.0/.dockerignore

@@ -0,0 +1,15 @@
+*.pyc
+*.pyd
+*.pyo
+.git
+.github/
+.gitignore
+.idea
+.mypy_cache
+.pytest_cache
+.ruff_cache
+.vscode
+__pycache__
+benchmarks/
+docs/
+tests/

kreuzberg-3.4.0/.github/workflows/publish-docker.yml

@@ -0,0 +1,101 @@
+# .github/workflows/publish-docker.yml
+
+name: Publish Docker Images
+
+on:
+  workflow_run:
+    workflows: ["Release"]
+    types:
+      - completed
+    branches:
+      - main
+
+jobs:
+  build-and-push:
+    runs-on: ubuntu-latest
+    if: ${{ github.event.workflow_run.conclusion == 'success' }}
+    permissions:
+      contents: read
+      packages: write
+
+    strategy:
+      matrix:
+        include:
+          - name: core
+            extras: ""
+            tag_suffix: "" # The base image tag (includes API + tesseract)
+          - name: easyocr
+            extras: "easyocr"
+            tag_suffix: "-easyocr"
+          - name: paddle
+            extras: "paddleocr"
+            tag_suffix: "-paddle"
+          - name: gmft
+            extras: "gmft"
+            tag_suffix: "-gmft"
+          - name: all
+            extras: "all"
+            tag_suffix: "-all"
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          ref: ${{ github.event.workflow_run.head_branch }}
+
+      - name: Get release version
+        id: get_version
+        run: |
+          echo "VERSION=${{ github.event.workflow_run.head_branch }}" >> $GITHUB_OUTPUT
+          # If triggered by a tag, extract version
+          if [[ "${{ github.event.workflow_run.head_branch }}" =~ ^v[0-9]+\.[0-9]+\.[0-9]+ ]]; then
+            echo "VERSION=${{ github.event.workflow_run.head_branch }}" >> $GITHUB_OUTPUT
+          else
+            # Get the latest tag
+            git fetch --tags
+            echo "VERSION=$(git describe --tags --abbrev=0)" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Log in to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Extract metadata (tags, labels) for Docker
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: goldziher/kreuzberg
+          tags: |
+            # Release version tag (e.g., v3.0.0-easyocr)
+            type=raw,value=${{ steps.get_version.outputs.VERSION }}${{ matrix.tag_suffix }}
+            # Latest tag for each variant (e.g., latest-easyocr)
+            type=raw,value=latest${{ matrix.tag_suffix }}
+
+      - name: Build and push Docker image
+        uses: docker/build-push-action@v5
+        with:
+          context: .
+          file: ./.docker/Dockerfile
+          platforms: linux/amd64,linux/arm64
+          push: true
+          build-args: |
+            EXTRAS=${{ matrix.extras }}
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+
+      - name: Update Docker Hub README
+        uses: peter-evans/dockerhub-description@v4
+        if: matrix.name == 'core'
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+          repository: goldziher/kreuzberg
+          readme-filepath: ./.docker/README.md

{kreuzberg-3.3.0 → kreuzberg-3.4.0}/.gitignore

@@ -6,28 +6,29 @@
 *.py[cod]
 *.suo
 *.user
-.DS_store
+*temp/
 .coverage
 .coverage*
+.cursorrules
 .dist/
+.DS_store
 .env
 .idea/
+.kreuzberg/
 .mypy_cache/
 .pytest_cache/
 .python-version
+.ropeproject
 .ruff_cache/
 .run/
 .venv/
 .vscode/
 .windsurfrules
-.cursorrules
-CLAUDE.md
-GEMINI.md
 __pycache__/
+benchmark_results.json
+CLAUDE.md
 coverage.xml
+docker-compose.yaml
+GEMINI.md
 prompt_template.egg-info/
 requirements.txt
-Dockerfile
-docker-compose.yaml
-benchmark_results.json
-.kreuzberg/

{kreuzberg-3.3.0 → kreuzberg-3.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kreuzberg
-Version: 3.3.0
+Version: 3.4.0
 Summary: A text extraction library supporting PDFs, images, office documents and more
 Project-URL: homepage, https://github.com/Goldziher/kreuzberg
 Author-email: Na'aman Hirschfeld <nhirschfed@gmail.com>
@@ -34,12 +34,15 @@ Provides-Extra: all
 Requires-Dist: click>=8.2.1; extra == 'all'
 Requires-Dist: easyocr>=1.7.2; extra == 'all'
 Requires-Dist: gmft>=0.4.2; extra == 'all'
+Requires-Dist: litestar[opentelemetry,standard,structlog]>=2.1.6; extra == 'all'
 Requires-Dist: paddleocr>=3.1.0; extra == 'all'
 Requires-Dist: paddlepaddle>=3.1.0; extra == 'all'
 Requires-Dist: rich>=14.0.0; extra == 'all'
 Requires-Dist: semantic-text-splitter>=0.27.0; extra == 'all'
 Requires-Dist: setuptools>=80.9.0; extra == 'all'
 Requires-Dist: tomli>=2.0.0; (python_version < '3.11') and extra == 'all'
+Provides-Extra: api
+Requires-Dist: litestar[opentelemetry,standard,structlog]>=2.1.6; extra == 'api'
 Provides-Extra: chunking
 Requires-Dist: semantic-text-splitter>=0.27.0; extra == 'chunking'
 Provides-Extra: cli
@@ -63,10 +66,14 @@ Description-Content-Type: text/markdown
 [![Documentation](https://img.shields.io/badge/docs-GitHub_Pages-blue)](https://goldziher.github.io/kreuzberg/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
-Kreuzberg is a Python library for text extraction from documents. It provides a unified interface for extracting text from PDFs, images, office documents, and more, with both async and sync APIs.
+Kreuzberg is a **high-performance** Python library for text extraction from documents. **Benchmarked as one of the fastest text extraction libraries available**, it provides a unified interface for extracting text from PDFs, images, office documents, and more, with both async and sync APIs optimized for speed and efficiency.
 
 ## Why Kreuzberg?
 
+- **🚀 Substantially Faster**: Extraction speeds that significantly outperform other text extraction libraries
+- **⚡ Unique Dual API**: The only framework supporting both sync and async APIs for maximum flexibility
+- **💾 Memory Efficient**: Lower memory footprint compared to competing libraries
+- **📊 Proven Performance**: [Comprehensive benchmarks](https://github.com/Goldziher/python-text-extraction-libs-benchmarks) demonstrate superior performance across formats
 - **Simple and Hassle-Free**: Clean API that just works, without complex configuration
 - **Local Processing**: No external API calls or cloud dependencies required
 - **Resource Efficient**: Lightweight processing without GPU requirements
@@ -85,6 +92,9 @@ pip install kreuzberg
 
 # Or install with CLI support
 pip install "kreuzberg[cli]"
+
+# Or install with API server
+pip install "kreuzberg[api]"
 ```
 
 Install pandoc:
@@ -134,6 +144,31 @@ async def main():
 asyncio.run(main())
 ```
 
+## Docker
+
+Docker images are available for easy deployment:
+
+```bash
+# Run the API server
+docker run -p 8000:8000 goldziher/kreuzberg:latest
+
+# Extract files via API
+curl -X POST http://localhost:8000/extract -F "data=@document.pdf"
+```
+
+See the [Docker documentation](https://goldziher.github.io/kreuzberg/user-guide/docker/) for more options.
+
+## REST API
+
+Run Kreuzberg as a REST API server:
+
+```bash
+pip install "kreuzberg[api]"
+litestar --app kreuzberg._api.main:app run
+```
+
+See the [API documentation](https://goldziher.github.io/kreuzberg/user-guide/api-server/) for endpoints and usage.
+
 ## Command Line Interface
 
 Kreuzberg includes a powerful CLI for processing documents from the command line:
@@ -208,7 +243,31 @@ For comparison and selection guidance, see the [OCR Backends](https://goldziher.
 
 ## Performance
 
-Kreuzberg offers both sync and async APIs. Choose the right one based on your use case:
+Kreuzberg delivers **exceptional performance** compared to other text extraction libraries:
+
+### 🏆 Competitive Benchmarks
+
+[Comprehensive benchmarks](https://github.com/Goldziher/python-text-extraction-libs-benchmarks) comparing Kreuzberg against other popular Python text extraction libraries show:
+
+- **Fastest Extraction**: Consistently fastest processing times across file formats
+- **Lowest Memory Usage**: Most memory-efficient text extraction solution
+- **100% Success Rate**: Reliable extraction across all tested document types
+- **Optimal for High-Throughput**: Designed for real-time, production applications
+
+### 💾 Installation Size Efficiency
+
+Kreuzberg delivers maximum performance with minimal overhead:
+
+1. **Kreuzberg**: 71.0 MB (20 deps) - Most lightweight
+1. **Unstructured**: 145.8 MB (54 deps) - Moderate footprint
+1. **MarkItDown**: 250.7 MB (25 deps) - ML inference overhead
+1. **Docling**: 1,031.9 MB (88 deps) - Full ML stack included
+
+**Kreuzberg is up to 14x smaller** than competing solutions while delivering superior performance.
+
+### ⚡ Sync vs Async Performance
+
+Kreuzberg is the only library offering both sync and async APIs. Choose based on your use case:
 
 | Operation              | Sync Time | Async Time | Async Advantage    |
 | ---------------------- | --------- | ---------- | ------------------ |
@@ -218,11 +277,7 @@ Kreuzberg offers both sync and async APIs. Choose the right one based on your us
 | OCR processing         | 0.4s      | 0.7s       | **✅ 1.7x faster** |
 | Batch operations       | 38.6s     | 8.5s       | **✅ 4.5x faster** |
 
-**Rule of thumb:**
-
-- Use **sync** for simple documents and CLI applications
-- Use **async** for complex PDFs, OCR, and batch processing
-- Use **batch operations** for multiple files
+**Rule of thumb:** Use async for complex documents, OCR, batch processing, and backend APIs.
 
 For detailed benchmarks and methodology, see our [Performance Documentation](https://goldziher.github.io/kreuzberg/advanced/performance/).
 

{kreuzberg-3.3.0 → kreuzberg-3.4.0}/README.md

@@ -5,10 +5,14 @@
 [![Documentation](https://img.shields.io/badge/docs-GitHub_Pages-blue)](https://goldziher.github.io/kreuzberg/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
-Kreuzberg is a Python library for text extraction from documents. It provides a unified interface for extracting text from PDFs, images, office documents, and more, with both async and sync APIs.
+Kreuzberg is a **high-performance** Python library for text extraction from documents. **Benchmarked as one of the fastest text extraction libraries available**, it provides a unified interface for extracting text from PDFs, images, office documents, and more, with both async and sync APIs optimized for speed and efficiency.
 
 ## Why Kreuzberg?
 
+- **🚀 Substantially Faster**: Extraction speeds that significantly outperform other text extraction libraries
+- **⚡ Unique Dual API**: The only framework supporting both sync and async APIs for maximum flexibility
+- **💾 Memory Efficient**: Lower memory footprint compared to competing libraries
+- **📊 Proven Performance**: [Comprehensive benchmarks](https://github.com/Goldziher/python-text-extraction-libs-benchmarks) demonstrate superior performance across formats
 - **Simple and Hassle-Free**: Clean API that just works, without complex configuration
 - **Local Processing**: No external API calls or cloud dependencies required
 - **Resource Efficient**: Lightweight processing without GPU requirements
@@ -27,6 +31,9 @@ pip install kreuzberg
 
 # Or install with CLI support
 pip install "kreuzberg[cli]"
+
+# Or install with API server
+pip install "kreuzberg[api]"
 ```
 
 Install pandoc:
@@ -76,6 +83,31 @@ async def main():
 asyncio.run(main())
 ```
 
+## Docker
+
+Docker images are available for easy deployment:
+
+```bash
+# Run the API server
+docker run -p 8000:8000 goldziher/kreuzberg:latest
+
+# Extract files via API
+curl -X POST http://localhost:8000/extract -F "data=@document.pdf"
+```
+
+See the [Docker documentation](https://goldziher.github.io/kreuzberg/user-guide/docker/) for more options.
+
+## REST API
+
+Run Kreuzberg as a REST API server:
+
+```bash
+pip install "kreuzberg[api]"
+litestar --app kreuzberg._api.main:app run
+```
+
+See the [API documentation](https://goldziher.github.io/kreuzberg/user-guide/api-server/) for endpoints and usage.
+
 ## Command Line Interface
 
 Kreuzberg includes a powerful CLI for processing documents from the command line:
@@ -150,7 +182,31 @@ For comparison and selection guidance, see the [OCR Backends](https://goldziher.
 
 ## Performance
 
-Kreuzberg offers both sync and async APIs. Choose the right one based on your use case:
+Kreuzberg delivers **exceptional performance** compared to other text extraction libraries:
+
+### 🏆 Competitive Benchmarks
+
+[Comprehensive benchmarks](https://github.com/Goldziher/python-text-extraction-libs-benchmarks) comparing Kreuzberg against other popular Python text extraction libraries show:
+
+- **Fastest Extraction**: Consistently fastest processing times across file formats
+- **Lowest Memory Usage**: Most memory-efficient text extraction solution
+- **100% Success Rate**: Reliable extraction across all tested document types
+- **Optimal for High-Throughput**: Designed for real-time, production applications
+
+### 💾 Installation Size Efficiency
+
+Kreuzberg delivers maximum performance with minimal overhead:
+
+1. **Kreuzberg**: 71.0 MB (20 deps) - Most lightweight
+1. **Unstructured**: 145.8 MB (54 deps) - Moderate footprint
+1. **MarkItDown**: 250.7 MB (25 deps) - ML inference overhead
+1. **Docling**: 1,031.9 MB (88 deps) - Full ML stack included
+
+**Kreuzberg is up to 14x smaller** than competing solutions while delivering superior performance.
+
+### ⚡ Sync vs Async Performance
+
+Kreuzberg is the only library offering both sync and async APIs. Choose based on your use case:
 
 | Operation              | Sync Time | Async Time | Async Advantage    |
 | ---------------------- | --------- | ---------- | ------------------ |
@@ -160,11 +216,7 @@ Kreuzberg offers both sync and async APIs. Choose the right one based on your us
 | OCR processing         | 0.4s      | 0.7s       | **✅ 1.7x faster** |
 | Batch operations       | 38.6s     | 8.5s       | **✅ 4.5x faster** |
 
-**Rule of thumb:**
-
-- Use **sync** for simple documents and CLI applications
-- Use **async** for complex PDFs, OCR, and batch processing
-- Use **batch operations** for multiple files
+**Rule of thumb:** Use async for complex documents, OCR, batch processing, and backend APIs.
 
 For detailed benchmarks and methodology, see our [Performance Documentation](https://goldziher.github.io/kreuzberg/advanced/performance/).
 

{kreuzberg-3.3.0 → kreuzberg-3.4.0}/docs/advanced/performance.md

@@ -4,18 +4,29 @@ Kreuzberg provides both synchronous and asynchronous APIs, each optimized for di
 
 ## Quick Reference
 
-| Use Case            | Recommended API              | Reason                       |
-| ------------------- | ---------------------------- | ---------------------------- |
-| CLI tools           | `extract_file_sync()`        | Lower overhead, simpler code |
-| Web applications    | `await extract_file()`       | Better concurrency           |
-| Simple documents    | `extract_file_sync()`        | Faster for small files       |
-| Complex PDFs        | `await extract_file()`       | Parallelized processing      |
-| Batch processing    | `await batch_extract_file()` | Concurrent execution         |
-| OCR-heavy workloads | `await extract_file()`       | Multiprocessing benefits     |
+| Use Case            | Recommended API              | Reason                                 |
+| ------------------- | ---------------------------- | -------------------------------------- |
+| CLI tools           | `extract_file_sync()`        | Lower overhead, simpler code           |
+| **Backend APIs**    | `await extract_file()`       | **Always use async in async contexts** |
+| Web applications    | `await extract_file()`       | Better concurrency                     |
+| Simple documents    | `extract_file_sync()`        | Faster for small files                 |
+| Complex PDFs        | `await extract_file()`       | Parallelized processing                |
+| Batch processing    | `await batch_extract_file()` | Concurrent execution                   |
+| OCR-heavy workloads | `await extract_file()`       | Multiprocessing benefits               |
 
-## Benchmark Results
+## Competitive Performance
 
-All benchmarks were conducted on macOS 15.5 with ARM64 (14 cores, 48GB RAM) using Python 3.13.3.
+[Comprehensive benchmarks](https://github.com/Goldziher/python-text-extraction-libs-benchmarks) comparing Kreuzberg against other popular Python text extraction libraries demonstrate:
+
+- **Fastest Extraction**: Consistently fastest processing times across file formats
+- **Lowest Memory Usage**: Most memory-efficient text extraction solution
+- **Smallest Installation**: 71.0 MB vs competitors ranging from 145.8 MB to 1,031.9 MB
+- **100% Success Rate**: Reliable extraction across all tested document types
+- **Optimal for High-Throughput**: Designed for real-time, production applications
+
+## Internal Benchmark Results
+
+All internal benchmarks were conducted on macOS 15.5 with ARM64 (14 cores, 48GB RAM) using Python 3.13.3.
 
 ### Single Document Processing
 
@@ -50,6 +61,29 @@ All benchmarks were conducted on macOS 15.5 with ARM64 (14 cores, 48GB RAM) usin
 1. **Simpler Path**: Direct execution without thread/process coordination
 1. **Fast Startup**: Immediate execution for quick operations
 
+### Backend API Considerations
+
+**Important**: When working in an async context (like FastAPI, Django async views, aiohttp), **always use the async API** even for simple documents:
+
+```python
+# ✅ Correct: Use async in async contexts
+async def extract_endpoint(file_path: str):
+    result = await extract_file(file_path)  # Non-blocking
+    return result
+
+# ❌ Wrong: Sync in async context blocks the event loop
+async def extract_endpoint(file_path: str):
+    result = extract_file_sync(file_path)  # Blocks event loop!
+    return result
+```
+
+**Why this matters:**
+
+- Sync operations in async contexts block the entire event loop
+- This prevents other requests from being processed concurrently
+- Backend throughput drops dramatically
+- Use async consistently throughout your async application stack
+
 ### The Crossover Point
 
 The performance crossover occurs around **10KB file size** or when **OCR is required**:
@@ -219,6 +253,17 @@ Choose your API based on your specific needs:
 
 - **Sync for simplicity**: CLI tools, simple documents, single-threaded applications
 - **Async for scale**: Web applications, batch processing, complex documents
+- **Async for backends**: **Always use async in async contexts** (FastAPI, Django async, etc.)
 - **Batch for efficiency**: Multiple files, concurrent processing requirements
 
+### Key Decision Points
+
+1. **Are you in an async context?** → Use async API
+1. **Processing multiple files?** → Use batch operations
+1. **Simple single document in sync context?** → Sync may be faster
+1. **Complex documents or OCR required?** → Use async API
+1. **Building a web API?** → Use async API
+
 The performance characteristics will vary based on your specific documents, hardware, and usage patterns. We recommend benchmarking with your actual data to make informed decisions.
+
+**Remember**: Kreuzberg is benchmarked as one of the fastest text extraction libraries available, delivering superior performance regardless of which API you choose.