LiteLLM (gateway)

Route LiteLLM requests through the Respan gateway to use 250+ models from different providers. Only your RESPAN_API_KEY is needed - no separate provider keys required.

Setup

Install packages

$ pip install litellm

Set environment variables

$ export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"

No provider key needed - the Respan gateway handles provider authentication.

Point LiteLLM to the Respan gateway

1 import os
2 import litellm
3 
4 response = litellm.completion(
5     api_key=os.environ["RESPAN_API_KEY"],
6     api_base=os.getenv("RESPAN_BASE_URL", "https://api.respan.ai/api"),
7     model="gpt-5.5",
8     messages=[{"role": "user", "content": "Hello!"}],
9 )
10 print(response.choices[0].message.content)

Switch models

Use another OpenAI model ID through the same gateway endpoint.

1 litellm.completion(api_key=..., api_base="https://api.respan.ai/api", model="gpt-5.5", messages=messages)
2 litellm.completion(api_key=..., api_base="https://api.respan.ai/api", model="gpt-5-mini", messages=messages)

LiteLLM can be forced onto its OpenAI transport for non-OpenAI models, but that makes the transport provider easy to confuse with the model provider. This page avoids showing Claude or Gemini as OpenAI-provider models; use the Respan API or OpenAI SDK gateway pages for provider-neutral Claude and Gemini examples.

See the full model list.

Respan parameters

Pass additional Respan parameters via extra_body for gateway features.

1 response = litellm.completion(
2     api_key=os.environ["RESPAN_API_KEY"],
3     api_base="https://api.respan.ai/api",
4     model="gpt-5.5",
5     messages=[{"role": "user", "content": "Hello!"}],
6     extra_body={
7         "customer_identifier": "user-123",
8         "metadata": {"session_id": "abc123"},
9         "thread_identifier": "conversation_456",
10     },
11 )

See Respan params & metadata for the full list.