🏆 LLM-Dispatcher: Intelligent LLM Routing

LLM-Dispatcher is an intelligent Python package that automatically selects the best Large Language Model (LLM) for your specific task based on performance metrics, cost optimization, and real-time availability.

🚀 Features

• 🧠 Intelligent Switching: Performance-based LLM selection using credible benchmarks
• 📊 Real Metrics: Based on 2024-2025 benchmark data (MMLU, HumanEval, GPQA, AIME, etc.)
• 💰 Cost Optimization: Dynamic cost-quality balancing
• 🔄 Smart Fallbacks: Automatic failover with intelligent routing
• 🎯 Multi-Modal: Support for text, vision, audio, and structured output
• ⚡ Real-time: Live performance monitoring and optimization
• 🔧 Simple API: Decorator-based usage with minimal configuration

📈 Supported Providers

Provider	Models	Capabilities
OpenAI	GPT-4, GPT-4 Turbo, GPT-3.5 Turbo	Text, Vision, Function Calling, Structured Output
Anthropic	Claude-3 Opus, Sonnet, Haiku	Text, Vision, Reasoning, Code
Google	Gemini 2.5 Pro, Flash, Ultra	Text, Vision, Multimodal, Fast
xAI	Grok 3 Beta	Text, Reasoning, Math

🎯 Performance Benchmarks

Based on latest 2024-2025 data:

Model	MMLU	HumanEval	GPQA	AIME	HellaSwag	ARC	VQA
GPT-4	86.3%	67.4%	82.1%	91.2%	95.1%	96.4%	78.2%
Claude-3 Opus	84.6%	67.4%	84.6%	89.8%	94.2%	95.8%	76.8%
Gemini 2.5 Pro	84.0%	65.2%	84.0%	87.3%	93.8%	94.2%	74.5%
Grok 3 Beta	82.1%	63.8%	84.6%	93.3%	92.1%	93.8%	71.2%

🚀 Quick Start

Installation

pip install llm-dispatcher

Basic Usage

from llm_dispatcher import llm_dispatcher

@llm_dispatcher
def generate_text(prompt: str) -> str:
    """Automatically routed to the best LLM for text generation."""
    return prompt

# Usage
result = generate_text("Write a story about a robot")
print(result)

Advanced Configuration

from llm_dispatcher import LLMSwitch, TaskType

# Initialize with custom configuration
switch = LLMSwitch(
    providers={
        "openai": {"api_key": "sk-..."},
        "anthropic": {"api_key": "sk-ant-..."},
        "google": {"api_key": "..."}
    },
    config={
        "prefer_cost_efficiency": True,
        "max_latency_ms": 2000,
        "fallback_enabled": True
    }
)

@switch.route(task_type=TaskType.CODE_GENERATION)
def generate_code(description: str) -> str:
    """Automatically uses the best model for code generation."""
    return description

@switch.route(task_type=TaskType.VISION_ANALYSIS)
def analyze_image(image_path: str) -> dict:
    """Automatically uses vision-capable models."""
    return {"analysis": "Image processed"}

📚 Documentation

• Installation Guide
• Quick Start Tutorial
• Advanced Configuration
• API Reference
• Benchmark Data
• Contributing

🧪 Development

Setup Development Environment

git clone https://github.com/ashhadahsan/llm-dispatcher.git
cd llm-dispatcher
pip install -e ".[dev]"
pre-commit install

Testing & Coverage

Run Tests

# Run all tests
pytest

# Run with coverage
pytest --cov=src/llm_dispatcher --cov-report=term-missing --cov-report=html

# Run specific test categories
pytest tests/test_core_switching.py -v
pytest tests/test_providers/ -v
pytest tests/test_multimodal.py -v

Current Test Coverage

• Overall Coverage: 47% (7,344 statements, 3,886 missed)
• Core Components: 98% coverage on base classes
• Provider Integration: 48-91% coverage across providers
• Multimodal Support: 47-86% coverage
• Monitoring & Analytics: 0% coverage (new features)

Coverage Reports

# Generate HTML coverage report
pytest --cov=src/llm_dispatcher --cov-report=html
open htmlcov/index.html

# Generate XML report for CI/CD
pytest --cov=src/llm_dispatcher --cov-report=xml

# Update coverage badge in README
python scripts/update_coverage_badge.py

Automated Coverage

• GitHub Actions: Automated coverage reporting on every PR
• Codecov Integration: Real-time coverage tracking
• Coverage Badge: Automatically updated in README
• Coverage Reports: Available as GitHub Actions artifacts

Run Benchmarks

pytest tests/benchmarks/ -v

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

Development Setup

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests
5. Submit a pull request

🧪 Testing

LLM-Dispatcher includes comprehensive tests to ensure reliability and performance.

Test Coverage

Current test coverage: 22% (29/29 core tests passing)

• ✅ Core Functionality: All basic tests passing
• ✅ Performance Tests: Latency and throughput tests working
• ✅ Benchmark Utils: Metric calculation and validation tests
• ✅ Provider Integration: OpenAI provider tests
• ✅ Configuration: Switch configuration and validation tests

Running Tests

# Install with dev dependencies
pip install -e ".[dev]"

# Run all tests
pytest

# Run with coverage
pytest --cov=src/llm_dispatcher --cov-report=html

# Run specific test categories
pytest tests/test_basic.py                    # Core functionality
pytest tests/test_performance.py              # Performance tests
pytest tests/test_benchmark_utils.py          # Benchmark utilities

Test Categories

Category	Tests	Status	Description
Core	27	✅ Passing	Basic functionality, configurations, providers
Performance	1	✅ Passing	Latency and throughput testing
Benchmarks	1	✅ Passing	Metric calculations and validation

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Benchmark data from Musaix, 51D.co, and Learnopoly
• LLM providers: OpenAI, Anthropic, Google, xAI
• Open source community

📞 Support

• 📧 Email: ashhadahsan@mail.com
• 🐛 Issues: GitHub Issues
• 💬 Discussions: GitHub Discussions

---

Built with ❤️ for the AI community