📊 Neatlogs - Observability Platform

Neatlogs is a comprehensive observability platform that provides detailed logging, monitoring, and analytics for LLM applications in production environments.

Using Neatlogs with LiteLLM

LiteLLM provides success_callbacks and failure_callbacks, allowing you to easily integrate Neatlogs for comprehensive tracing and monitoring of your LLM operations.

Integration

Use just a few lines of code to instantly log your LLM responses across all providers with Neatlogs:

import litellm

# Configure LiteLLM to use Neatlogs
litellm.success_callback = ["neatlogs"]

# Make your LLM calls as usual
response = litellm.completion(
    model="gpt-3.5-turbo",
    messages=[{"role": "user", "content": "Hello, how are you?"}],
)

Complete Code Example

import os
from litellm import completion

# Set environment variables
os.environ["OPENAI_API_KEY"] = "your-openai-key"
os.environ["NEATLOGS_API_KEY"] = "your-neatlogs-api-key"

# Configure LiteLLM to use Neatlogs
litellm.success_callback = ["neatlogs"]

# Make LLM call
response = completion(
    model="gpt-4",
    messages=[{"role": "user", "content": "Hi 👋 - I'm using Neatlogs!"}],
)

print(response)

Configuration Options

The Neatlogs integration can be configured through environment variables:

• NEATLOGS_API_KEY (str, required): Your Neatlogs API key

Advanced Usage

Async Support

Neatlogs fully supports async operations:

import asyncio
import litellm

litellm.success_callback = ["neatlogs"]

async def main():
    response = await litellm.acompletion(
        model="gpt-4",
        messages=[{"role": "user", "content": "Hello async world!"}],
    )
    print(response)

asyncio.run(main())

Streaming Support

Neatlogs automatically handles streaming responses:

import litellm

litellm.success_callback = ["neatlogs"]

response = litellm.completion(
    model="gpt-4",
    messages=[{"role": "user", "content": "Tell me a story"}],
    stream=True
)

for chunk in response:
    print(chunk.choices[0].delta.content, end="")

Error Tracking

Neatlogs also tracks failed requests:

import litellm

# Enable both success and failure callbacks
litellm.success_callback = ["neatlogs"]
litellm.failure_callback = ["neatlogs"]

try:
    response = litellm.completion(
        model="gpt-4",
        messages=[{"role": "user", "content": "Hello!"}],
        # This might fail due to invalid parameters
        temperature=3.0  # Invalid temperature
    )
except Exception as e:
    print(f"Error: {e}")
    # Error will be logged to Neatlogs automatically

Data Tracked

Neatlogs captures comprehensive data for each LLM request:

• Request Details: Model, provider, messages, parameters
• Response Data: Completion text, token usage, cost
• Performance Metrics: Response time, latency
• Error Information: Failure reasons and stack traces
• Metadata:session IDs

Supported Models and Providers

Neatlogs works with all LiteLLM-supported providers:

• OpenAI (GPT-3.5, GPT-4, etc.)
• Anthropic (Claude)
• Google (Gemini)
• Azure OpenAI
• AWS Bedrock
• And 100+ more providers

Support

For issues or questions, please refer to:

• Neatlogs Documentation
• Github