How to Use Make.com with DeepSeek Qwen

In this step-by-step guide, you will learn how to connect DeepSeek Qwen to make.com so you can build powerful AI automations, multi-agent workflows, and cost-efficient scenarios without writing complex code.

The instructions below are based on the official DeepSeek Qwen tutorial from the Make team and walk you through setup, configuration, and practical usage patterns.

Why Combine Make.com and DeepSeek Qwen?

DeepSeek Qwen is an open-source, enterprise-grade language model family designed for low latency, low cost, and long context processing. When integrated with make.com, it becomes a flexible AI engine you can plug into any automation.

Using this combination, you can:

• Run AI text and code generation inside your scenarios.
• Orchestrate multi-agent workflows for complex tasks.
• Access long context windows for documents and conversations.
• Cut inference costs compared with many other models.

All these benefits are available through a standard API-style connection, which you can configure once and reuse in multiple make.com scenarios.

Prerequisites for Make.com and DeepSeek Qwen

Before you start, make sure you have the following ready:

• An active make.com account with access to the scenario editor.
• A DeepSeek Qwen account and API key (from DeepSeek or a supported provider).
• Basic familiarity with building scenarios and working with modules in make.com.
• A clear use case, such as content generation, document chat, or tool-using agents.

Step 1: Get Your DeepSeek Qwen API Key

The first step is to obtain an API key so that make.com can securely communicate with DeepSeek Qwen.

1. Sign in to your DeepSeek Qwen or supported host dashboard.

2. Navigate to the developer or API section.

3. Create a new API key dedicated to your make.com projects.

4. Copy the key and store it in a safe place; you will need it during connection setup.

Do not share this key publicly, and consider creating different keys for testing and production scenarios.

Step 2: Create a New Scenario in Make.com

Next, prepare a scenario where you will add DeepSeek Qwen as an AI component.

1. Log in to your make.com dashboard.

2. Click Create a new scenario.

3. Add your trigger module (for example, a webhook, a scheduler, or an app event such as a form submission).

4. Save the scenario and give it a descriptive name like “DeepSeek Qwen Content Generator”.

This scenario will become the workspace where you configure prompts, agents, and downstream actions.

Step 3: Connect DeepSeek Qwen to Make.com

Once the scenario exists, you can add and authenticate the DeepSeek Qwen module.

1. Inside the scenario editor, click the plus icon to add a new module.

2. Search for the DeepSeek Qwen or OpenAI-compatible module, depending on how the model is exposed by the source page.

3. Select the action corresponding to generating text, running chat, or using an agent.

4. When prompted for a connection, click Add or Create a connection.

5. Paste your API key, configure the base URL if required by your provider, and confirm.

After the connection is saved, you can reuse it across multiple modules and scenarios on make.com.

Step 4: Choose the Right DeepSeek Qwen Model

DeepSeek Qwen offers several specialized models. According to the source article at the official DeepSeek Qwen guide, you can choose from families optimized for general reasoning, coding, mobile use, and more.

Typical options include:

• Qwen general models for chat, content creation, and reasoning.
• Qwen code-oriented models for programming help and code generation.
• Lightweight models for fast and cheap inference on frequent tasks.

Inside the DeepSeek Qwen module in make.com, select the model name that best fits your use case. If you are not sure, start with the recommended general-purpose model described in the source article and adjust later based on results.

Step 5: Configure Your Prompt and Inputs in Make.com

To get useful output, you need a clear prompt and a good structure for inputs and variables.

Set Base Instructions in Make.com

Define what the model should do overall:

• Open the DeepSeek Qwen module.

• In the System or Instruction field, add high-level guidelines (for example: “You are a helpful support assistant that answers concisely and clearly.”).

• Use text variables from previous modules (such as form fields or webhook data) to make the AI behavior dynamic.

Use Dynamic User Prompts

Then, configure the user message or prompt:

• Map text from earlier modules into the prompt field, such as questions from a customer, task descriptions, or document snippets.

• Add context like language preferences, tone, or formatting instructions.

• Optionally include IDs or metadata so downstream modules can trace the AI response back to a specific request.

By carefully structuring instructions in make.com, you help DeepSeek Qwen perform consistently on every run.

Step 6: Build Multi-Agent Workflows on Make.com

The article describes how DeepSeek Qwen supports agent-style workflows, where one or more AI agents can call tools, browse data, or coordinate tasks. You can orchestrate this behavior with branches and routers in make.com.

Design Agent Roles

Create different roles in your scenario, for example:

• A Planner agent that breaks a complex request into steps.
• A Research agent that reads documents or external APIs.
• A Writer agent that produces final output for the user.

Each role can be represented by its own DeepSeek Qwen module with tailored system prompts and settings inside make.com.

Route Data Between Agents

Connect the agents in sequence using tools available in make.com:

• Routers to branch flows based on conditions.
• Text and array functions to combine or split content.
• Iterators to loop through lists of tasks or documents.

This pattern lets you build robust multi-agent processes while keeping each prompt focused and simple.

Step 7: Control Cost and Performance

The DeepSeek Qwen family is designed to be cost-effective, but you should still configure your make.com scenario to avoid unnecessary usage.

• Use smaller models when exact reasoning is not required.
• Limit max tokens for responses to what you truly need in downstream modules.
• Summarize inputs before passing them to the model if they are very long.
• Cache or store results when repeated queries would otherwise hit the model multiple times.

Experiment with temperature and other parameters in the DeepSeek Qwen module to balance creativity with determinism.

Step 8: Test, Monitor, and Iterate in Make.com

After configuration, you should thoroughly test your scenario before moving it into production.

1. Run the scenario with sample data and inspect the output from each DeepSeek Qwen module.

2. Use the execution log in make.com to see how inputs, prompts, and responses flow through the modules.

3. Refine prompts, adjust parameters, or change models based on observed behavior.

4. Set up error handling paths for failed calls or timeouts.

Continuous iteration will gradually improve quality, reliability, and cost efficiency.

Advanced Tips for Make.com and DeepSeek Qwen

Once the basic integration works, you can expand your automations with more advanced patterns inspired by the source guide.

• Document chat: Store documents in a database or third-party tool, retrieve relevant excerpts, and feed them to DeepSeek Qwen via make.com to answer questions.
• Tool calling: Let the model plan actions while make.com executes them using built-in app modules and APIs.
• Hybrid workflows: Combine traditional rule-based logic with AI reasoning to keep critical steps fully deterministic.

If you need expert implementation support, you can also consult automation specialists such as Consultevo, who work extensively with AI and make.com systems.

Conclusion: Build Scalable AI Automations on Make.com

By integrating DeepSeek Qwen with make.com, you can create fast, affordable AI workflows that power content creation, support, research, and complex multi-agent tasks. You only need to configure the connection once, then reuse it across many scenarios to keep your automations consistent and maintainable.

Use the steps in this how-to guide—getting an API key, creating a scenario, connecting DeepSeek Qwen, designing prompts, building agents, and optimizing cost—to turn your ideas into production-ready AI automation flows.