> ## Documentation Index > Fetch the complete documentation index at: https://docs.acedata.cloud/llms.txt > Use this file to discover all available pages before exploring further. # Use AceDataCloud in n8n > Import a ready-to-run chat workflow using n8n's built-in HTTP Request node. The [sample workflow](https://raw.githubusercontent.com/AceDataCloud/Docs/main/examples/n8n/acedatacloud-chat.json) uses n8n's built-in Manual Trigger, Edit Fields, and HTTP Request nodes. It sends one non-streaming request to AceDataCloud Chat Completions and exposes the answer, response ID, model, and token counts. You do not need a community node. ## Native nodes and model entry Use independent service nodes after a trigger and search by service name before building HTTP requests. Cloud verification and self-hosted npm installation are separate. * [Suno](/en/integrations/n8n-suno) * [Seedance](/en/integrations/n8n-seedance) * [HappyHorse](/en/integrations/n8n-happyhorse) * [GPT Image](/en/integrations/n8n-gpt-image) * [Nano Banana](/en/integrations/n8n-nano-banana) For chat, install [Ace Data Cloud Chat Model](https://github.com/AceDataCloud/AceChatN8N) and connect it to an Agent/Basic LLM Chain **Model** input, with no Base URL field. It is not a default-installed provider or verified Cloud listing. Published 0.1.2 provides native **Model Name or ID** catalog choices and exact-ID expressions; reading the list does not generate a paid answer. The HTTP example below is an optional low-level recipe, not a replacement for these native nodes or official distribution. ## Import and run 1. Follow [Get an API key](#get-an-api-key) below, then check the [current model price](https://platform.acedata.cloud/models?utm_source=opensource\&utm_medium=referral\&utm_campaign=evergreen\&utm_content=n8n_docs_en_pricing) and available balance. 2. In n8n, create a workflow. Choose **Import from URL** and paste the [workflow JSON URL](https://raw.githubusercontent.com/AceDataCloud/Docs/main/examples/n8n/acedatacloud-chat.json). You can also download the JSON and choose **Import from File**. 3. Open **AceDataCloud chat**. **Authentication → Generic Credential Type** and **Generic Auth Type → Bearer Auth** are already set. Click **Connect to Bearer Auth**, then create or select an HTTP Bearer Auth credential. Enter the token **without** a `Bearer ` prefix. Select the credential and save the workflow before running it. Credentials are not included in the JSON. 4. In **Set prompt and model**, edit the prompt and, if needed, replace `gpt-4o-mini` with another model supported by [Chat Completions](/en/guides/openai/openai_chat_completions). The sample limits output to 80 tokens. Adjust `max_tokens` in the HTTP Request body if your task needs a longer answer. 5. Click **Execute workflow** once. Open **Read answer and usage** to inspect `answer`, `response_id`, `model`, and token counts. The workflow starts inactive and runs only when you click it. Keep the token in the n8n credential, not in an Edit Fields node or the workflow JSON. Review n8n's execution-data retention settings before processing sensitive prompts. ## Get an API key 1. Open [Applications](https://platform.acedata.cloud/console/applications?utm_source=opensource\&utm_medium=referral\&utm_campaign=evergreen\&utm_content=n8n_docs_en_api_key) and sign in or register. If **General application** is missing, choose **Initialize General Application**. 2. Copy the application's **API Key**, or choose **Keys / Manage Keys → Create** to create a separately named key for n8n. A general application's key can use services available to your account; a dedicated application's key is limited to its service. 3. Check **Expiration Time**, **Set Usage Limitation**, and **Allowed APIs**. Allow Chat Completions and the model you intend to use. Check your balance before running the workflow; any trial allowance is whatever your account currently displays. 4. Copy the raw token into n8n's HTTP Bearer Auth credential. Do not include `Bearer `, quotation marks, or spaces. A `platform-...` account-management token cannot replace a generation API key. ## Connect your automation Replace **Start a manual test** with your own trigger and map its text to `prompt` in **Set prompt and model**. The HTTP Request node posts to `https://api.acedata.cloud/openai/chat/completions` with `stream: false`. Downstream nodes can use `answer` and `response_id` from **Read answer and usage**. Each execution can consume credits. The final node's token counts do not establish the amount charged: check the AceDataCloud usage record for the actual charge. Start with one manual run before adding a schedule or webhook. Monitor successful and failed n8n executions, duration, response IDs, token counts, and the matching AceDataCloud usage records. ## Handle errors and retries The HTTP Request node fails the execution on an unsuccessful HTTP response. Open that node to inspect the status and error body: | Status or symptom | Next step | | - | - | | `400` | Check the model, prompt, and request body. | | `401` or `403` | Check the selected token and its access. Correct the credential before another run. | | `429` | Check the rate limit and wait before retrying the failed execution. | | Timeout or `5xx` | Check the AceDataCloud usage record and response ID, if one was returned, before retrying. The first request may have completed even if n8n did not receive its response. | Automatic retries are disabled for this POST because replaying a completed request could incur a second charge. If you need retries, add them only after deciding how to reconcile ambiguous outcomes in your workflow. This JSON is available for import directly. It will appear in n8n's built-in **Templates** search only after n8n publishes a separate template listing. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.