Mirmer AI Python SDK

A Python client library for the Mirmer AI multi-LLM consultation system.

Installation

pip install mirmer-ai

Quick Start

from mirmer import Client

# Initialize client with API key
client = Client(api_key="your-api-key")

# Or use environment variable MIRMER_API_KEY
client = Client()

# Send a query and get council response
response = client.query("What is the meaning of life?")

# Access the three stages
print("Stage 1 - Individual Responses:")
for model_response in response.stage1:
    print(f"{model_response.model}: {model_response.response}")

print("\nStage 2 - Peer Rankings:")
for ranking in response.stage2:
    print(f"{ranking.model}: {ranking.parsed_ranking}")

print("\nStage 3 - Chairman Synthesis:")
print(response.stage3.response)

Features

• 3-Stage Council Process: Query multiple AI models, get peer reviews, and receive synthesized consensus
• Streaming Support: Real-time updates as each stage completes
• Async/Await: Full async support with AsyncClient
• Conversation Management: Create, list, search, and delete conversations
• Usage Tracking: Monitor your API consumption and limits
• Type Safety: Complete type hints for IDE autocomplete and type checking
• Error Handling: Comprehensive exception hierarchy for graceful error handling

Usage Examples

Streaming Responses

from mirmer import Client

client = Client(api_key="your-api-key")

# Stream council process updates in real-time
for update in client.stream("Explain quantum computing"):
    if update.type == "stage1_complete":
        print(f"Stage 1 complete: {len(update.data['stage1'])} responses")
    elif update.type == "stage3_complete":
        print(f"Final answer: {update.data['response']}")

Async Usage

import asyncio
from mirmer import AsyncClient

async def main():
    async with AsyncClient(api_key="your-api-key") as client:
        response = await client.query("What is machine learning?")
        print(response.stage3.response)

asyncio.run(main())

Conversation Management

from mirmer import Client

client = Client(api_key="your-api-key")

# Create a new conversation
conversation = client.create_conversation(title="AI Discussion")

# Add messages to the conversation
response = client.query("What is AI?", conversation_id=conversation.id)

# List all conversations
conversations = client.list_conversations()

# Search conversations
results = client.search_conversations("machine learning")

# Get specific conversation
conv = client.get_conversation(conversation.id)

# Delete conversation
client.delete_conversation(conversation.id)

Usage Statistics

from mirmer import Client

client = Client(api_key="your-api-key")

# Check your usage
usage = client.get_usage()
print(f"Used: {usage.queries_used_today}/{usage.daily_limit}")
print(f"Tier: {usage.tier}")

Configuration

from mirmer import Client

client = Client(
    api_key="your-api-key",
    base_url="https://api.mirmer.ai",  # Custom API endpoint
    timeout=60.0,                       # Request timeout in seconds
    max_retries=3                       # Max retry attempts for failed requests
)

Error Handling

from mirmer import Client, AuthenticationError, RateLimitError, APIError

client = Client(api_key="your-api-key")

try:
    response = client.query("Hello")
except AuthenticationError:
    print("Invalid API key")
except RateLimitError as e:
    print(f"Rate limit exceeded. Reset at: {e.reset_time}")
except APIError as e:
    print(f"API error: {e.message} (status: {e.status_code})")

Requirements

• Python 3.8+
• httpx >= 0.24.0
• pydantic >= 2.0.0
• python-dateutil >= 2.8.0

License

MIT License - see LICENSE file for details

Support

• Documentation: https://docs.mirmer.ai
• Issues: https://github.com/mirmer-ai/python-sdk/issues
• Email: support@mirmer.ai