> ## Documentation Index
> Fetch the complete documentation index at: https://laminar.sh/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# LiteLLM observability
> Trace LiteLLM calls to any model it routes to with one integration: prompts, token usage, cost, and latency in Laminar's transcript view.
## Overview
[LiteLLM](https://www.litellm.ai/) is a framework/library for building LLM applications that simplifies accessing many models across different providers.
## Default configuration
```sh theme={null}
pip install -U 'lmnr[all]'
```
Laminar auto-instruments LiteLLM. No callback is required.
```python theme={null}
import os
import litellm
from lmnr import Laminar
Laminar.initialize(project_api_key=os.environ["LMNR_PROJECT_API_KEY"])
```
`LaminarLiteLLMCallback` is deprecated and does nothing. Remove it if you already use it.
Example code:
```python theme={null}
import os
import litellm
from lmnr import Laminar
Laminar.initialize(project_api_key=os.environ["LMNR_PROJECT_API_KEY"])
response = litellm.completion(
model="gpt-5-nano",
messages=[
{"role": "user", "content": "What is the capital of France?"}
],
)
```
## Related integrations
Trace the OpenAI SDK directly in TypeScript and Python.
Trace the Anthropic SDK directly in TypeScript and Python.
Trace OpenRouter requests with model and cost attribution.
Trace Cohere chat, embed, and rerank calls.
Trace LangChain chains and LangGraph graphs.
Browse every provider, framework, coding agent, and browser integration Laminar supports.
