| #https://www.notion.so/n8n/Frontmatter-432c2b8dff1f43d4b1c8d20075510fe4 | |
| contentType: howto | |
| title: Call an API to fetch data | |
| description: Use the n8n workflow tool to load data from an API using the HTTP Request node into your AI workflow. | |
| # Call an API to fetch data | |
| Use n8n to bring data from any [API](/glossary.md#api) to your AI. This workflow uses the [Chat Trigger](/integrations/builtin/core-nodes/n8n-nodes-langchain.chattrigger/index.md) to provide the chat interface, and the [Call n8n Workflow Tool](/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.toolworkflow.md) to call a second workflow that calls the API. The second workflow uses AI functionality to refine the API request based on the user's query. | |
| [[ workflowDemo("file:///advanced-ai/examples/let_your_ai_call_an_api.json") ]] | |
| ## Key features | |
| This workflow uses: | |
| * [Chat Trigger](/integrations/builtin/core-nodes/n8n-nodes-langchain.chattrigger/index.md): start your workflow and respond to user chat interactions. The node provides a customizable chat interface. | |
| * [Agent](/integrations/builtin/cluster-nodes/root-nodes/n8n-nodes-langchain.agent/index.md): the key piece of the AI workflow. The Agent interacts with other components of the workflow and makes decisions about what tools to use. | |
| * [Call n8n Workflow Tool](/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.toolworkflow.md): plug in n8n workflows as custom tools. In AI, a tool is an interface the AI can use to interact with the world (in this case, the data provided by your workflow). The AI model uses the tool to access information beyond its built-in dataset. | |
| * A [Basic LLM Chain](/integrations/builtin/cluster-nodes/root-nodes/n8n-nodes-langchain.chainllm.md) with an [Auto-fixing Output Parser](/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.outputparserautofixing.md) and [Structured Output Parser](/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.outputparserstructured/index.md) to read the user's query and set parameters for the API call based on the user input. | |
| ## Using the example | |
| --8<-- "_snippets/examples-color-key.md" | |