> For the complete documentation index, see [llms.txt](https://docs.diaflow.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.diaflow.io/workflow-builder/nodes/public-ai-llm-models/deepseek-cloud.md). # Deepseek Cloud The **Deepseek Cloud** node lets you use DeepSeek inside a Diaflow workflow with your own API key. It is a good fit when your team wants strong text generation and analysis while keeping close control over provider choice and usage cost. You can use this node for tasks such as writing drafts, summarizing content, answering business questions, and turning source material into short, usable outputs. {% hint style="info" %} Diaflow runs the workflow. DeepSeek charges usage to your own DeepSeek account. {% endhint %} ## What this node does The node takes your instruction, sends it to the selected DeepSeek model, and returns the result to the next step in your workflow. In most business workflows, that result is text. You can also pass content from earlier steps into this node. For example, you can send customer messages, internal notes, uploaded text, or retrieved knowledge and ask DeepSeek to turn that content into a useful business answer. ## When to use it Use **Deepseek Cloud** when you want Diaflow to work with your own DeepSeek setup. This is useful when your company wants a flexible text model provider and needs direct control over billing and model selection. Common use cases include: * Summarize reports, notes, or long text into key takeaways. * Draft replies, internal updates, or first-pass business content. * Turn raw input into clean categories, decisions, or structured answers. ## Before you start Make sure you have: * An active DeepSeek account. * A valid DeepSeek API key. * Permission to use the model you want. ## How to set it up Set up the node in this order: 1. In **Credentials**, choose your saved DeepSeek key. 2. In **Model**, choose the DeepSeek model that fits your task. 3. In **Prompt**, describe what you want the model to do. 4. Run the node and review the result.
If you want the model to work from earlier workflow data, insert that data into the prompt with `@`. ## What each field means ### Credentials This is where you select the DeepSeek API key your team wants to use. If the correct key is not listed, add it first, then refresh the list if needed. ### Model This is where you choose the DeepSeek model. The list depends on the key and account you connected. For most business users, the choice is simple: * Choose a general text model for writing, summaries, and question answering. * Choose a stronger model when output quality matters more than speed or cost. * Test with a real business task before using it widely. If you are unsure, start with a general-purpose model and compare the result using a real example from your workflow. ### Prompt This is the instruction you give the model. Write it in plain language and focus on the result you want. A strong prompt usually includes: * the task * the context * the expected format For example: ``` Summarize this internal project update for senior managers. Keep the message concise and business-focused. Return 3 key points and 2 recommended next actions. ``` If your workflow already has data from previous steps, insert that data into the prompt with `@` so the model uses that content directly. ## Advanced configurations Open **Show advanced configurations** when you need more control. ### Enable caching Caching reuses a previous result when the same input runs again. This can help save time and reduce repeated calls. Use caching when the same request is likely to run more than once and the answer does not need to change every time. ### Caching time This controls how long the cached result stays available. Use a shorter time when the source content changes often. Use a longer time when the same request repeats across multiple runs. ## What you get back The node returns the model’s result to the next step in your workflow. In most cases, that result is text. You can send it to an output node, store it, transform it, or pass it into another business step. ## Tips for business users * Start with one clear task per prompt. * Ask for a specific format such as bullets, short summary, or structured text. * Test with real business content before using the node in production. * If the answer is too broad, make the prompt narrower and more specific. ### Next steps * Compare other providers in [Public AI/LLM Models](/workflow-builder/nodes/public-ai-llm-models.md). * Learn the workflow basics in [Overview](/workflow-builder/overview.md). * Browse more nodes in [Component List](/workflow-builder/component-list.md). --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.diaflow.io/workflow-builder/nodes/public-ai-llm-models/deepseek-cloud.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.