How To Use DeepSeek? | DeepSeek Tutorial For Beginners | Getting Started With DeepSeek | Simplilearn

[music] Deep Seek has become one of the most talked about tools because people are not just using it for random questions anymore. Students are using it to understand difficult topics, working professionals are using it to write emails and summarize notes, content creators are using it to generate ideas and developers are using it to build small practical tools. In 2026, Deep Seek has become even more relevant with the launch of Deep Seek V4 preview, which is now available on web, app and API. The new version focuses on stronger reasoning, better coding support and longer context, which means it can handle larger prompts, longer documents and more detailed problem-solving tasks. Deep Seek's official API update also mentions two new versions, which is Deep Seek V4 Pro for more powerful tasks and Deep Seek V4 Flash for faster and lighter usage. So, in simple words, Deep Seek is a tool where you type a question, task or instruction and it helps you create explanations, summaries, emails, study plans, code, research points and much more. But, the important thing is this, just opening Deep Seek and typing a basic question is not enough. To get real value from it, you need to know how to ask it clearly, how to use it for practical tasks, how to check its answers and how to connect it with tools like VS Code for coding-based work. So, this topic is important because Deep Seek is becoming useful for learners, job seekers, professionals and developers. But, it should also be used carefully. For example, Reuters has reported that Deep Seek has faced privacy and security in different regions. So, users should avoid entering passwords, confidential files, private documents or sensitive company data. So, in this tutorial, we will understand Deep Seek in a simple and practical way and see how it can help us in real work and learn how to use it safely and confidently. So, here's what we will learn in today's course. What is Deep Seek? We will understand Deep Seek in simple beginner-friendly language. Deep Seek versus ChatGPT. Here, we will compare both tools and understand when to use Deepseek. How to access Deepseek. Here we will see how to use it through website, app, and API. Better prompts. Here we will learn how to clear questions and get better answers. Daily work and research. We will use Deepseek for emails, notes, summaries, and learning plans. File analysis. Then we will see how to summarize and organize long text files. Coding with Deepseek. We will use it to explain code, fix errors, and create simple programs. VS Code demo. Then we will connect Deepseek with VS Code and Python to generate simple code. Followed by running it practically, fixing errors, and building small tools like calculator, to-do lists, expense trackers, or study planners. Mini project. We will then create a simple personal work assistant. And finally, we will learn how to use Deepseek carefully and revise the key points. So, before we move on, let me share something really exciting with you guys. If today's Deepseek tutorial made you realize how useful these tools can be for writing, coding, research, and real workplace tasks, then you should definitely check out Advance Executive Program in Applied Generative AI in collaboration with Simplilearn. So, this program is designed for learners who do not want to stop at just using tools casually, but want to understand how these technologies are applied for real business situations. You get to learn important skills like Python, prompt writing, building useful applications, working with modern tools like Azure AI Studio, Microsoft Copilot, ChatGPT, Hugging Face, Streamlit, Gradio, and more. So, the program also includes hands-on projects where you work on practical use cases like building an HR assistant, creating a logo design, developing a news assistant, analyzing customer orders, building a Python game, and creating business insights from data. So, instead of learning theory, you get to practice how these tools can solve real problems in areas like business, content, customer support, data analysis, and productivity. The best part is that this program also includes live classes, expert-led learning, integrated labs, master classes, capstone projects, and career support, which makes it useful for students, working professionals, managers, developers, and anyone who wants to build strong future-ready skills. So, if you want to move from simply using tools like Deep Seek to actually building practical solutions with them, do check out Advanced Executive Program in Applied Generative AI in collaboration with Simplilearn. So, the link is in the description box. Go check it out. Before we move on, here is a quick quiz question. What is the best way to get better answers from Deep Seek? Is it A, ask one-word questions, B, give clear instructions, C, share private passwords, or is it D, avoid checking answers? Let us know in the comment section below. Hello, everyone, and welcome back. In this coding demo, we are going to build practical tools using Deep Seek and Python. We will start from the setup and then slowly build small useful programs like a chat program, a prompt tester, email writer, meeting note summarizer, study plan generator, file summarizer, code debugging helper, and finally a small personal work assistant. Now, before we write any code, let us first prepare our system. So, for this demo, we need three things. First, we need Python, VS Code, and a Deep Seek API key. Python will help us write and run the code. VS Code would be our code editor, and the Deep Seek API key will allow our Python program to connect with Deep Seek. So, Deep Seek's official API documentation says that an API can be used with the Open API SDK by setting up the base URL to Deep Seek's API endpoint. And the newer [snorts] model names include Deep Seek V4 Flash and Deep Seek V4 Pro. And the official docs also mention that the older names like Deep Seek Chat and Deep Seek Reasoner are planned to be retired. So, after July 24, 2026, we will be using the newer model name. So, now let's start with the installation. First, install Python from the official Python website. So, while installing Python on Windows, make sure that the option called add Python to path is selected. So, this helps you run Python from the terminal. So, after installing Python, open command prompt or the terminal and type Python version. So, as you can see in my system Python has been installed properly and it shows the Python version. Now, let's go ahead and install VS Code. So, VS Code is where we will write all our project files. So, once VS Code is installed, open it and create one new folder called Deep Seek Coding Demo. So, since I've already created the directory, I will go ahead and open it. So, I will use the command CD Deep Seek Coding Demo. So, as you can see this has opened VS Code for us. So, now this creates a new project folder. So, we move inside the folder and this opened VS Code. Now, let's go ahead and create a virtual environment. So, virtual environment is like a separate workspace for this project. So, it keeps all of the project packages inside one place. For that, let's go ahead and type Python followed by m v n v space {dot} v n v. So, now we have to go ahead and activate it. So, for Windows, use this particular line of code. So, that's {dot} v n v {slash} scripts and {slash} activate. So, as you can see it has been activated. So, once it has been activated, we will go ahead and install the required packages. So, let's go ahead and install it. So, pip install open API Python space {dot} e n v. So, here open API helps our Python code connect with Deep Seek because Deep Seek supports an open AI compatible format. So, Python {dot} e n v helps us keep our API key safely inside a separate file. So, now create a new file called {dot} e n v. {dot} e n v. So, now to go ahead and create a new file, type the following command. So, it's new followed by item. So, it says that the file here already exists because I had already created it earlier. So, now to open and edit it in notepad, open {dot} e n v. So, now go ahead and add your API key like this. So, that will be Deep Seek. So, as you can see here, here I've pasted my actual Deep Seek API key. So, we're not writing the key directly inside the Python file because it's better to keep private keys separate. So, now go ahead and create a file called deepseek_client.py. So, as you can see here, I've created a file called deepseek_client.py. So, within this you go ahead and type the following code. So, first we start off with import os, then from .env import .env from .env import open. Okay. So, now let's go ahead and understand this code in simple words. So, first we import the required packages, then we load the .env file so that Python can read the API key. After that, we check whether the API key is available. So, if the key is missing, the program shows a clear error. Then we create a client. So, this client is a connection between our Python project and DeepSeek. Finally, we create a reusable function called ask_deepseek. So, this function accepts a user message and sends it to DeepSeek and returns the answer. This file is important because we reuse it in upcoming demos. So, now that our setup is ready, let us move on and build our first DeepSeek chat program. So, now that we have created our connection file, let's build our first simple program. So, in this program, we will ask one question and print the answer in the terminal. So, create a new file called first_chat.py. So, to create any new file, what you will have to do is you will have to go create a new file, and as we are typing our code, you go right click on the same file, and then move on save, and then you can rename your file as you wish. So, I will name this as first chat.py. So, now within this I will go ahead and type the following code. So, that's from deepseek_client import then ask_deepseek. So, now let's go ahead and understand this code. So, the first line imports the ask_deepseek function from the file that we created earlier. Then we store one question inside a variable called question. So, after that, we send the question to DeepSeek and store the answer inside our answer variable. And finally, we train both the question and the answer. So, now let's go ahead and run the program. So, you move back to your terminal and type Python followed by the file name, which is first chat.py. So, once we run it, DeepSeek will respond with a beginner-friendly explanation of Python. So, this is the simplest DeepSeek API program, and when one question goes in, another answer comes out. So, now that we've built our first chat program, let's move on to understand how better instructions can give better answers. So, now while running this program, you may have seen an error like insufficient balance. As you can see the error over here. So, this means that our API key is valid, but the API account doesn't have enough balance to complete the request. So, now in a real project, the solution is simple. You need to top up your API account and then rerun the program. But now for this tutorial, we will not go ahead with the top up right now. So, instead, I will show you a mock demo. So, a mock demo here means that we will temporarily return a sample response from our Python function without actually calling the live DeepSeek API. So, this is useful when you want to test your code flow, record a demo, or explain how the program works without spending API credits. But remember, in a real application, you should use your own valid API key with active billing or balance. So, just for now, let's temporarily update our DeepSeek client file. So, replace only the ask DeepSeek function with this particular function. Then let's go ahead and continue with the script. So, let's try to run the file again. So, that's Python first client first chat.py. So, this time, instead of sending the question to a live DeepSeek API, our function returns a sample response. So, the program flow is the same. The question goes into the ask deep function, and the function gives back an answer. And first chat.py prints that answer in the terminal. So, this helps us understand the complete working of the program even without using live API balance. So, later when you have API balance available, you can replace this mock function with the real API code again. So now that we can ask questions, let us learn how to improve the quality of the answers. So a prompt simply means the instruction we give to Deepseek. So if we give a very general instruction, the answer will also have to be general. But if we give a clear instruction with the details, the answer can become more useful. So now let's go ahead and create a file called prompt tester.py. So I have done this already. Now from there, go ahead and type the following code. So that's from Deepseek client and then we import then ask Deepseek. Then let's go ahead and type week prompt. So now that that's done, let us go and understand what's happening here. So we have created two prompts as you can see. So the first one is short and general and it only says tell me about data analytics. Because this prompt is not very specific and the answer may not be too broad. The second prompt here is much cleaner. So it tells Deepseek who the learner is, what type of language to use, what points to include and what style to follow. Now let's go ahead and run the program. So go back to your terminal and type Python followed by prompt tester. So as you can see when we compare both outputs, we will notice that the stronger output gives a much better and more useful answer. So the lesson here is simple. Don't just ask short questions, give context, mention the audience and explain the output format that you want. So now that we are better, let us create a simple email writer. So this program will generate a professional email based on the purpose that we provide. So create a new file called email writer.py. I've already done this over here. And type the following code. So that's from Deepseek import _deepseek email purpose. So now let's go ahead and understand the code. We're taking three inputs from the user. So first, we ask for the purpose of the email, then we ask who the email is for and then we ask what tone should the email have. After that, we place three values inside a clear prompt. So the prompt asks Deepseek to write a professional email with a subject line. So, now let's go ahead and run the program. Python email. So, for example, we can write requesting 2 days of leave. Who is the email for? It can be for your manager. Enter the tone. Can be polite. So, this program will generate a professional leave request email. So, this is useful for students, working professionals, job seekers, and anyone who wants to write cure emails quickly. So, now that we've created an email writer, let's build a meeting notes summarizer. So, many times meeting notes are messy. People write quick points, deadlines, names, and tasks. So, this program will convert rough notes into a clean summary and action items. So, create a new file called meeting summarizer.py. And in the following file, write the following code. That's from Deep Seek client. Then we import ask Deep Seek. So, now that we've typed the code, let's go ahead and understand this. So, first we have rough meeting notes inside a multi-line text. So, these notes include people, tasks, deadlines, and one possible risk. Then we send these notes to Deep Seek and ask it to organize them into a clean format. We clearly mention the output format, which is summary, action items, owners, deadlines, risks, and next step. So, now let's go ahead and run the program. Python meeting sum- marizer. So, as you can see, after we run the program, the output is much organized than the original notes. So, this demo is very useful because it shows how Deep Seek can help with real office work. So, instead of manually cleaning meeting notes, we can use simple programs to organize them. So, now that you have seen a work-related use case, let us build something useful for learners. So, this program will create a study plan for any topic. So, let's go ahead and create a new file called study plan generator.py. And from there, from Deep Seek client, then import ask Deep Seek. So, now let's go ahead and understand this code. So, we ask the user for three things, which is the topic, the number of days, and the learner's level. Then, we build a prompt using these values. For example, if the user enters Python, 7 days, and beginner, the program will create a 7-day beginner-friendly Python learning plan. So, now let's go ahead and run this program. Python, followed by study plan generator. So, here let's go ahead and enter the topic as Python. How many days do you have? That'll be 7 days, as I said. And enter your level. I will go ahead and type beginner. So, as you can see, this program is helpful for students, freshers, and professionals who want to learn a new skill with clear plan. So, as you can see, the study plan has been generated. So, now that we've created a study plan generator, let's move on to file summarization. So, now that we've created a study plan generator, let's build a file summarizer. So, in this demo, we will read content from a text file and then ask DeepSeek to summarize it. So, first create a file called sample_article.text, within which you have to type the following prompt. Now, let's go ahead and create a file called file_summarizer.py. So, let's go ahead and type the code for this. From DeepSeek. So, now let's go ahead and understand this code. So, the open function opens the text file. The read function reads the full content from the file, and then we send that content to DeepSeek and ask it for a simple summary. So, this is very useful when we have long notes, reports, articles, or study material. So, now let's go ahead and run this program. So, Python file_summarizer. So, as you can see, this program will read the file and generate a simple summary. So, now that we know how to summarize files, let's move on to a coding-related use case, which is debugging. So, now that we have worked with files, let's build a code debugging helper. So, debugging means finding and fixing errors in code. So, this program will take broken Python code, send it to DeepSeek, and ask for a simple explanation and corrected version. So, create a new file called code_debugging_helper, and from there I type this code. From deepseek and import ask deepseek. So, now let's go ahead and try to understand this particular code. So, this code is trying to calculate the average of marks, but there is a problem here. The code uses the variable count, but count was never created. So, Python will not know what count means. So, instead of directly giving the fixed code, we ask deepseek to explain the error, why it happened, how to fix it, and how to give the correct code. So, error, why it happens, how to fix it, and then give the corrected code. So, now let's go ahead and run the program. So, when we run the file, this is what we get. Python code debugging helper .py. Yeah. So, as you can see, the output here explain the mistake in simple English. So, now let's go ahead and create the corrected program separately. So, here we've created a file called average calculator.py and add the following lines of code. So, once we're done with that, we'll go ahead and run it. So, we have average calculator.py. So, let's go back and explain the code. So, here, len numbers counts how many values are inside the list. And then, we divide the total marks by the count. That gives us the average. So, now that we have built a debugging helper, let's create a simple command line chatbot. So, now that we have seen a single task program, let's build a simple chatbot that keeps asking questions until we stop. So, here we've created a file called simple chatbot.py. Add the following lines of code. So, let me go ahead and explain this particular code. So, the program first prints a welcome message, and then it starts a loop using while true. So, this means the program will continue to run until we stop. And inside the loop, the user types a question. And if the user types exit, the program stops. Otherwise, the question is sent to deepseek, and the answer is printed. So, now let's go ahead and run this. So, that's Python simple chatbot.py. So, now we can ask questions like explain HTML in simple words or give me five project ideas for Python beginners or write a professional LinkedIn post about learning data analytics. So let me go ahead and ask what is Python. So as you can see this is a basic chatbot and it answers one question at a time. But there is one limitation. It does not remember the previous messages. So now that we have built a simple chatbot, let's upgrade it with a conversation history. So now that we have created a basic chatbot, let's go ahead and improve this. So in the previous chatbot each question was treated separately. So now we will create a chatbot that remembers the current conversation while the program is running. So let's go ahead and create a file called chatbot with history and type the following code in it. So now let's try to understand this particular code. So we create a list called messages and this list stores a full conversation. So first it stores the system message which tells the chatbot how to behave and then every time the user asks a question we will add the question to the list. After Deep Seek replies we will also add the answer to the same list. So this way the chatbot can understand the current question along with the earlier messages. So now let's go ahead and run this program. So we have Python chatbot with history. So first let's go ahead and ask I am a beginner learning Python. So now that we have built a chatbot with conversation history, let's learn how to get a clean output in JSON format. So JSON is a simple format that programs can easily read. So it is useful when we want to use the answer in a fixed structure like lists, summaries, deadlines, names or categories. So Deep Seek's official JSON output guide says that we should set response format to type JSON object, include the word JSON in the prompt and provide an example of the JSON format that we want. So let's go ahead and create a file called JSON output demo.py and type the following code. So now for the explanation first we import JSON because we want to work with the JSON output. Then we give some meeting notes. After that we give Deep Seek a clear system message and ask it to return only valid JSON. Then we use JSON loads to convert the response into a Python object. Finally, we print the summary and tasks in a clean format. So, let's go ahead and run Python JSON output demo.py. So, this is useful when we want to build real tools because programs work better with structured data rather than long paragraphs. So, now that we know how to get clean JSON output, let's add safety checks and error handling. So, now that we have created multiple tools, let's make sure that our program is safer. So, good program should not be sending private information by mistake. It's also supposed to handle errors properly. So, for example, if the API key is missing, internet is not working, or the service is unavailable, the program should show a clear message. So, let's go ahead and create a new file called safe question helper.py. Then type the following code. So, let's go ahead and understand this. So, we created a list called blocked words. So, these are the words that may indicate private or sensitive information. Then we create a function called is safe question. So, this function checks whether the user's question can contain a blocked word or not. So, if the question contains sensitive words, then the program stops and asks the user to remove private details. We also use try and accept. So, this program helps handle errors properly. So, if something goes wrong, the program may show helpful messages instead of just crashing down suddenly. So, now let's go ahead and run this program. So, that's Python safe question helper.py. So, when we go ahead and run this code, and when we type our question, what is Python? So, we get an answer as follows. Now, try entering your password. So, now if we ask the question, my password is 1 2 3 4 5 6, can you remember it? So, if we go ahead and run this, so, as you can see, this added safety and error handling to the same, and we got the output as please remove sensitive details and try again. So, now that we have added safety and error handling, let's bring everything together into one small mini project. So, now that we have built many small demos, let's combine them into one final project. So, this final project would be a personal work assistant. So, it will show us a menu and allow the users to choose what they want to. So, the user can choose email writing, meeting notes summarizing, study planning, file summarizing, or code debugging help. So, create a new file called personal work assistant.py and type the following code. So, we have the code over here. We created separate functions for separate tasks. So, the email write function creates emails, the summarize meeting notes function summarizes the notes, the create study plan function here creates a learning plan, and the summarize file function reads the file and then summarizes it. As well as the debug code function helps explain and fix code. Then, we created a menu using a while true loop. So, the user chooses an option and then the program runs matching functions. So, this is better than keeping everything in one large block because functions make the code cleaner and easier to understand. So, now let's run the final project. So, let's run the code. So, that's Python personal work assistant. py. So, here we have the option of picking from any of the options provided. So, I'll go with one. So, for now, I'll be showing you a mock demo of what the actual predicted outputs will be all in one file. Here, then we can enter the purpose of the email, which will be leave manager polite. So, the final project here shows how Deep Seek can be used in practical daily tasks with Python. So, now that we have completed the final mini project, let's move ahead and conclude. And that brings us to the end of this Deep Seek tutorial. We learned what Deep Seek is, why it's important, and how to use it for daily work, research files, coding, and how to connect it with Python inside VS Code. So, the key takeaway is simple. Deep Seek becomes more useful when you ask it clearly, review the output, and use it safely. Thanks for watching and subscribe to Simply Learn for more insights, and check out our related videos. Mhm.