LiteLLM | Respan Docs

LiteLLM provides a unified Python interface for calling 100+ LLM providers using the OpenAI format. Respan gives you full observability over every LiteLLM completion across providers — and gateway routing through the OpenAI-compatible Respan endpoint.

Set up Respan

Create an account at platform.respan.ai and grab an API key. For gateway, also add credits or a provider key.

Run npx @respan/cli setup to set up with your coding agent.

Example projects

Python examples

Tracing

Gateway

Setup

Install packages

$ pip install respan-ai respan-exporter-litellm litellm

Set environment variables

$ export OPENAI_API_KEY="YOUR_OPENAI_API_KEY"
$ export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"

OPENAI_API_KEY (or any provider key) is used for LLM requests. RESPAN_API_KEY is used to export traces to Respan.

Initialize and run

Register the Respan callback to log all completions automatically. Requests go directly to providers; the logs are sent to Respan.

1 import litellm
2 from respan_exporter_litellm import RespanLiteLLMCallback
3 
4 litellm.callbacks = [RespanLiteLLMCallback()]
5 
6 response = litellm.completion(
7     model="gpt-4.1-nano",
8     messages=[{"role": "user", "content": "Say hello in three languages."}],
9 )
10 print(response.choices[0].message.content)

View your trace

Open the Traces page to see your LiteLLM completions across providers as auto-traced spans.

Configuration

Parameter	Type	Default	Description
`api_key`	`str`	`RESPAN_API_KEY` env var	Respan API key.
`base_url`	`str \| None`	`None`	API base URL.

See the LiteLLM Exporter SDK reference for the full API.

Attributes

Pass Respan parameters inside metadata.respan_params on each call.

1 response = litellm.completion(
2     model="gpt-4.1-nano",
3     messages=[{"role": "user", "content": "Hello!"}],
4     metadata={
5         "respan_params": {
6             "workflow_name": "simple_logging",
7             "span_name": "single_log",
8             "customer_identifier": "user-123",
9             "thread_identifier": "conv_abc_123",
10             "metadata": {"plan": "pro"},
11         }
12     },
13 )

Attribute	Description
`customer_identifier`	Identifies the end user in Respan analytics.
`thread_identifier`	Groups related messages into a conversation.
`metadata`	Custom key-value pairs attached to the span.
`workflow_name`	Logical workflow grouping the call belongs to.
`span_name`	Custom name for the resulting span.

Async usage

The callback supports async completions automatically.

1 import litellm
2 from respan_exporter_litellm import RespanLiteLLMCallback
3 
4 litellm.callbacks = [RespanLiteLLMCallback()]
5 
6 response = await litellm.acompletion(
7     model="gpt-4.1-nano",
8     messages=[{"role": "user", "content": "Tell me a joke"}],
9 )

Multiple providers

LiteLLM’s unified interface means all providers are logged with the same callback.

1 import litellm
2 from respan_exporter_litellm import RespanLiteLLMCallback
3 
4 litellm.callbacks = [RespanLiteLLMCallback()]
5 
6 litellm.completion(model="gpt-4.1-nano", messages=[...])
7 litellm.completion(model="claude-sonnet-4-5-20250929", messages=[...])
8 litellm.completion(model="together_ai/meta-llama/Llama-3-70b", messages=[...])