DeepSeek

DeepSeek is a family of AI models developed by DeepSeek Artificial Intelligence Co., Ltd., which specializes in natural language processing (NLP) and machine learning.

This is a step-by-step guide on installing and configuring DeepSeek for ONLYOFFICE Docs.

Connecting and configuring DeepSeek models

  1. Follow this guide for general configuration.
  2. Once the plugin is installed, add AI models. Go to the AI tab, and open Settings.
  3. Select Edit AI models in the bottom left corner of the window and click Add.
  4. In the window that appears, select DeepSeek from the drop-down list.
    DeepSeek AI: dropdown listDeepSeek AI: dropdown list
  5. Input your API key in the provided field. The API key is issued by the AI provider and can be generated on the DeepSeek official website. Click the Update models list button.
  6. Select either DeepSeek Chat or DeepSeek Reasoner, or both if necessary, from the drop-down list:
    • DeepSeek Chat is an AI model designed for interactive conversations, offering fluid and context-aware dialogue capabilities. It provides intelligent responses with contextual memory, making it suitable for customer support, brainstorming sessions, and real-time assistance in document creation.
    • DeepSeek Reasoner is specialized in logical inference, structured reasoning, and complex problem-solving. It is particularly useful for users who need deep analytical insights, mathematical reasoning, or structured decision support.

      You can also connect other DeepSeek models through the Together AI provider.

    • Deepseek-ai/DeepSeek-R1 is a versatile, general-purpose AI model optimized for everyday office tasks such as drafting emails, generating reports, and summarizing documents. Prioritizes consistent quality over niche specialization. Unlike Chat (focused on interactive dialogue) and Reasoner (logical/mathematical problem-solving), R1 emphasizes versatility over analytical depth.
    • Deepseek-ai/DeepSeek-V3 is an advanced AI for complex, context-rich tasks requiring detailed responses, such as high-quality content generation, nuanced language processing, or in-depth data insights. V3 focuses on depth and detail in static outputs.
  7. In the row of icons, select the model used for: text, images, embeddings, audio processing, content moderation, real-time tasks, coding help, and visual analysis.
    DeepSeek AI: using modelsDeepSeek AI: using models
  8. Click OK and close the window. The model will appear in the list of available models in the AI model list.
  9. Assign different tasks to the AI. From the drop-down list, select the installed AI models that will perform these tasks. For example, DeepSeek Chat for the chatbot.
    DeepSeek AI: task assignmentDeepSeek AI: task assignment
  10. Access the AI features through the AI tab or by selecting text and right-clicking to open the context menu.
    DeepSeek AI: AI featuresDeepSeek AI: AI features

Host ONLYOFFICE Docs on your own server or use it in the cloud

Article with the tag:
Browse all tags