# Integrations

Anything that works with OpenAI's API and lets you change the base URL works with Continuum. These are the settings for popular tools.

In every case:

```text
Base URL   https://api.continuumlabs.ai/v1
API key    cl-your-key
Model      glm-5.2, or any ID from the models page
```

## OpenAI Agents SDK

```python
from agents import Agent, Runner, OpenAIChatCompletionsModel, set_tracing_disabled
from openai import AsyncOpenAI

client = AsyncOpenAI(base_url="https://api.continuumlabs.ai/v1", api_key="cl-your-key")
set_tracing_disabled(True)  # tracing would send data to OpenAI

agent = Agent(
    name="Assistant",
    instructions="Answer in Australian English.",
    model=OpenAIChatCompletionsModel(model="glm-5.2", openai_client=client),
)
print(Runner.run_sync(agent, "What is the GST rate in Australia?").final_output)
```

> Turn tracing off, as shown. Otherwise the SDK sends traces of your runs to OpenAI, outside Australia.

## LangChain

Install `langchain-openai`.

```python
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(base_url="https://api.continuumlabs.ai/v1", api_key="cl-your-key", model="glm-5.2", temperature=0)
print(llm.invoke("What is the GST rate in Australia?").content)
```

## LlamaIndex

Install `llama-index-llms-openai-like`.

```python
from llama_index.llms.openai_like import OpenAILike

llm = OpenAILike(api_base="https://api.continuumlabs.ai/v1", api_key="cl-your-key", model="glm-5.2", is_chat_model=True)
print(llm.complete("What is the GST rate in Australia?"))
```

## Vercel AI SDK

Install `ai` and `@ai-sdk/openai-compatible`.

```typescript
import { createOpenAICompatible } from '@ai-sdk/openai-compatible'
import { generateText } from 'ai'

const continuum = createOpenAICompatible({
  name: 'continuum',
  baseURL: 'https://api.continuumlabs.ai/v1',
  apiKey: process.env.CONTINUUM_API_KEY,
})

const { text } = await generateText({ model: continuum('glm-5.2'), prompt: 'What is the GST rate in Australia?' })
```

## n8n, Make and Zapier

Use the OpenAI node or module and create a credential with your Continuum key. Where the credential has a **Base URL** field, set it to `https://api.continuumlabs.ai/v1`. Then type the model ID in the model field.

If a tool has no base URL setting, it cannot connect to Continuum yet. Its HTTP request step, sending a `POST` to `https://api.continuumlabs.ai/v1/chat/completions`, always works.

## Coding assistants

Most coding assistants, including Cursor, Continue and Cline, accept an OpenAI-compatible provider. In the assistant's model settings, add a custom OpenAI-compatible model with the base URL, your key and the model ID.

To have an assistant write code against this API, give it [https://www.continuumlabs.ai/llms.txt](https://www.continuumlabs.ai/llms.txt).

## Something not working?

Tell us which tool and what you tried, at [hello@continuumlabs.ai](mailto:hello@continuumlabs.ai). We will add it here.