> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://respan.ai/docs/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://respan.ai/docs/_mcp/server.

# Pipecat (gateway)

> Set up Pipecat with the Respan gateway.

Route Pipecat's underlying LLM calls through the Respan gateway to use 1000+ models from different providers. Only your `RESPAN_API_KEY` is needed — no separate provider keys required.

## Setup

#### Install packages

```bash
pip install "pipecat-ai[openai]"
```

#### Set environment variables

```bash
export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"
export RESPAN_BASE_URL="https://api.respan.ai/api"
export RESPAN_MODEL="gpt-5.5"
```

No `OPENAI_API_KEY` needed — the Respan gateway handles provider authentication.

#### Point Pipecat to the Respan gateway

```python
import os

from pipecat.services.openai.llm import OpenAILLMService

llm = OpenAILLMService(
    api_key=os.environ["RESPAN_API_KEY"],
    base_url=os.getenv("RESPAN_BASE_URL", "https://api.respan.ai/api"),
    settings=OpenAILLMService.Settings(model=os.getenv("RESPAN_MODEL", "gpt-5.5")),
)
```

Then build and run your pipeline as usual.

## Switch models

Change the model setting on `OpenAILLMService` to use another OpenAI model through the same gateway-backed endpoint.

```python
OpenAILLMService(
    api_key=os.environ["RESPAN_API_KEY"],
    base_url="https://api.respan.ai/api",
    settings=OpenAILLMService.Settings(model="gpt-5.5"),
)

OpenAILLMService(
    api_key=os.environ["RESPAN_API_KEY"],
    base_url="https://api.respan.ai/api",
    settings=OpenAILLMService.Settings(model="gpt-5-mini"),
)
```

> **Note**
>
> `OpenAILLMService` is Pipecat's OpenAI-compatible service. This page avoids showing Claude or Gemini inside that OpenAI-named service; use the Respan API or OpenAI SDK gateway pages for provider-neutral Claude and Gemini examples.

See the [full model list](https://platform.respan.ai/platform/models).