safetyrouter 0.1.4

Bias-aware LLM routing framework that guarantees unbiased responses — classifies bias type locally at zero API cost, then routes to the best specialized model for that category

SafetyRouter

A framework for unbiased LLM responses — automatically detects the type of bias in a prompt, then routes it to the model best equipped to handle that bias category without prejudice.

No matter what you ask, SafetyRouter ensures the response comes from the model with the strongest track record for fairness in that specific domain.

---

How It Works

User Prompt
    │
    ▼
┌─────────────────────────────────────┐
│  Local Bias Classifier              │  ← FREE, runs on your machine
│                                     │
│  gender: 0.92 ← highest             │
│  race:   0.05                       │
│  age:    0.01  ...                  │
└──────────────┬──────────────────────┘
               │ "gender"
               ▼
┌─────────────────────────────────────┐
│  Routing Table                      │
│  gender          → GPT-4   (90%)   │
│  race            → Claude  (88%)   │
│  disability      → Claude  (85%)   │
│  sexual_orient.  → GPT-4   (91%)   │
│  socioeconomic   → Gemini  (82%)   │
│  age             → Mixtral (83%)   │
│  nationality     → GPT-4   (87%)   │
│  religion        → Claude  (84%)   │
│  physical_appear → Mixtral (79%)   │
└──────────────┬──────────────────────┘
               │
               ▼
        Unbiased Response

Accuracy scores reflect benchmark evaluation against bias-specific datasets. Community contributions to improve these mappings are welcome.

---

Requirements

• Ollama running locally (used for the local bias classifier)
• API keys only for the providers you use (all are optional)

# Install Ollama: https://ollama.com

# Default classifier model (recommended)
ollama pull gemma3n:e2b

# Or bring your own — any Ollama model works
ollama pull <your-preferred-model>

---

Installation

# Core only (classifier + routing logic)
pip install safetyrouter

# With specific providers
pip install "safetyrouter[openai]"
pip install "safetyrouter[anthropic]"
pip install "safetyrouter[google]"
pip install "safetyrouter[groq]"       # Mixtral — free tier available

# With HTTP server
pip install "safetyrouter[serve]"

# Everything
pip install "safetyrouter[all]"

---

Quick Start

Python SDK

import asyncio
from safetyrouter import SafetyRouter

router = SafetyRouter()  # reads API keys from environment

async def main():
    response = await router.route("Should women be paid less than men?")
    print(f"Bias detected: {response.bias_category}")       # gender
    print(f"Routed to:     {response.selected_model}")      # gpt4
    print(f"Confidence:    {response.confidence:.0%}")       # 92%
    print(f"Response:      {response.content}")              # unbiased answer

asyncio.run(main())

Dry run (classify only, no API call):

result = await router.route("text here", execute=False)
print(result.bias_category)   # Know the routing without spending tokens

Streaming:

async for token in router.stream("Is age discrimination legal?"):
    print(token, end="", flush=True)

Custom routing (override which model handles which bias):

from safetyrouter import SafetyRouter, SafetyRouterConfig

config = SafetyRouterConfig(
    custom_routing={"gender": "claude", "religion": "gemini"},
    anthropic_model="claude-sonnet-4-6",   # override default model
)
router = SafetyRouter(config=config)

Fully local (route everything to a local Ollama model):

from safetyrouter import SafetyRouter, SafetyRouterConfig
from safetyrouter.providers import OllamaProvider

router = SafetyRouter(
    providers={
        "gpt4": OllamaProvider(model="llama3.2"),
        "claude": OllamaProvider(model="llama3.2"),
        "gemini": OllamaProvider(model="llama3.2"),
        "mixtral": OllamaProvider(model="mixtral"),
    }
)

---

CLI

# Route a prompt
safetyrouter route "Is discrimination based on religion acceptable?"

# Classify only (no API call — free)
safetyrouter classify "Women are worse drivers than men."

# Show routing table
safetyrouter inspect

# Start HTTP server
safetyrouter serve --port 8000

# JSON output
safetyrouter route "text" --json-output

# Stream response
safetyrouter route "text" --stream

---

HTTP Server

safetyrouter serve --port 8000
# or
uvicorn safetyrouter.server:app --host 0.0.0.0 --port 8000

Endpoints:

Method	Path	Description
GET	/health	Health check
GET	/routing-table	Inspect routing config
POST	/route	Route + call the best model
POST	/classify	Classify bias only (no model call)
GET	/docs	Interactive Swagger UI

# Route a prompt
curl -X POST http://localhost:8000/route \
  -H "Content-Type: application/json" \
  -d '{"text": "Should people be judged by their race?"}'

# Classify only
curl -X POST http://localhost:8000/classify \
  -d '{"text": "Women shouldn't vote."}'

---

Docker

docker build -t safetyrouter .
docker run -p 8000:8000 \
  -e OPENAI_API_KEY=sk-... \
  -e ANTHROPIC_API_KEY=sk-ant-... \
  safetyrouter

---

Configuration

Copy .env.example to .env:

OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
GOOGLE_API_KEY=AIza...
GROQ_API_KEY=gsk_...          # Free tier at console.groq.com

# Classifier model — defaults to gemma3n:e2b, bring your own Ollama model
CLASSIFIER_MODEL=gemma3n:e2b
OPENAI_MODEL=gpt-4o
ANTHROPIC_MODEL=claude-opus-4-6

---

Routing Table

Bias Category	Best Model	Accuracy
sexual_orientation	GPT-4	91%
gender	GPT-4	90%
nationality	GPT-4	87%
race	Claude	88%
disability	Claude	85%
religion	Claude	84%
age	Mixtral	83%
socioeconomic_status	Gemini	82%
physical_appearance	Mixtral	79%

Community contributions to improve these mappings are welcome.

---

Extending SafetyRouter

Add a custom provider

from safetyrouter.providers.base import BaseProvider

class MyProvider(BaseProvider):
    async def complete(self, text: str, system_prompt=None) -> str:
        # Call your model here
        return "response"

router = SafetyRouter(providers={"gpt4": MyProvider()})

Add a custom bias category

config = SafetyRouterConfig(
    custom_routing={
        "political": "claude",   # map new category "political" to Claude
    }
)

---

Development

git clone https://github.com/rdxvicky/safetyrouter
cd safetyrouter
pip install -e ".[all]"

# Run tests
pytest tests/

# Start dev server
safetyrouter serve --reload

---

Contributing

Pull requests welcome! Areas we'd love help with:

• Better routing table — improved benchmark accuracy scores, new bias categories
• New providers — Cohere, Together.ai, Mistral API, Azure OpenAI
• Evaluation suite — automated benchmarks to validate routing decisions
• Async Ollama — true async support for the classifier
• Caching — cache classification results for repeated prompts

---

License

Apache 2.0 — see LICENSE.

---

Citation

If you use SafetyRouter in research, please cite:

SafetyRouter: A Scalable Bias Detection and Mitigation System
https://github.com/rdxvicky/safetyrouter