# Use BigModel.io with n8n

> Connect n8n to BigModel.io with an OpenAI credential whose Base URL is https://bigmodel.io/v1, then use GLM, Kimi or DeepSeek in the OpenAI Chat Model node by ID.

Web version: https://bigmodel.io/docs/n8n · Updated 2026-10-18

n8n's built-in **OpenAI** credential accepts a custom Base URL, so the **OpenAI Chat Model** node and AI Agent workflows can call BigModel.io models.

| Setting | Value |
|---|---|
| Credential type | OpenAI |
| API Key | your BigModel.io API key (`sk-...`) |
| Base URL | `https://bigmodel.io/v1` (must end in `/v1`) |
| Model (in the node) | **By ID**: a model ID from the [model catalog](https://bigmodel.io/models), e.g. `glm-5.3` |
| Use Responses API | off (Chat Completions) |

## Create the credential

1. In n8n, open **Credentials → Add credential** and choose **OpenAI**.
2. Paste your key from [API Keys](https://bigmodel.io/keys) into **API Key**.
3. Set **Base URL** to `https://bigmodel.io/v1`.
4. Save. n8n tests the credential by calling `GET https://bigmodel.io/v1/models`.

The Base URL is set on the credential. Recent versions of the OpenAI Chat Model node do not show a node-level base URL.

## Use it in a workflow

1. Add an **OpenAI Chat Model** node (for example under an AI Agent or a Basic LLM Chain) and select the credential.
2. For **Model**, choose **By ID** and enter the exact ID, for example `glm-5.3`. **From list** also works when the credential test lists your models.
3. Keep **Use Responses API** off. Chat Completions-only GLM and DeepSeek groups return 404 on `/v1/responses`.

## Verify the setup

Run the workflow once with a short prompt such as "ping". The node output should contain a reply, and the request should appear in [Usage](https://bigmodel.io/usage) with the model ID you entered.

## Troubleshooting

**Credential test fails with 403 `INSUFFICIENT_BALANCE`.** The test calls `/v1/models`, which needs a balance above $0. Top up in [Billing](https://bigmodel.io/purchase) and test again.

**Credential test fails with 404.** The Base URL is missing `/v1` or has an extra path. Use `https://bigmodel.io/v1`.

**401 `INVALID_API_KEY`.** The key has spaces or was deleted. Copy it again from [API Keys](https://bigmodel.io/keys).

**404 `not_found_error`: `... /v1/responses is not available`.** **Use Responses API** is on with a key from a Chat Completions-only group. Turn it off.

**503 `api_error`: `No available accounts: no available accounts supporting model: ...`.** The model ID is misspelled or not served by your key's group.

**429 during batch runs.** `rate_limit_error` means too many parallel requests ("Too many pending requests"); add a wait or lower batch size. `rate_limit_exceeded` with a message like "api key 5小时限额已用完" means a key spend window is used up.

## Related

- [Dify](https://bigmodel.io/docs/dify) for LLM apps and workflows.
- [Per-key limits & budgets](https://bigmodel.io/docs/key-limits) to cap what an automation can spend.
- [Rate limits](https://bigmodel.io/docs/rate-limits) for concurrency and 429 errors.