> For the complete documentation index, see [llms.txt](https://scade.gitbook.io/scade-knowledge-base/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://scade.gitbook.io/scade-knowledge-base/flow-examples/building-a-flow-video-transcription-and-summarization.md).
# Building a flow: video transcription and summarization
Life is too short for watching long videos? Let's build a workflow to transcribe and summarize any Youtube video.
1. Go to the [Flow](https://app.scade.pro/flow) and click the "Create" button.
2. It's a good practice to store all your input data in a single node. Start by typing `user input` in the search bar of the left panel. Locate **User-Defined Input Form** and drag it to the workspace.
3. This node has no pre-configured fields, so we should set them ourselves. Dive into the node’s settings and select **Configure Fields**.
4. Add a field to store the link to the video. Let's name this field `video_url`. We don't have to change the field type or anything else. Click **Save**.\
5. Paste the Youtube link to the node, hit the **Execute** button and then click **Save and execute** to run this node and generate the output. We need it to move further.\
\
A video that is used as an example in this flow:
\
\
Here's the output
6. Start typing in the search bar what you want to do next: `video transcription`. We've found two models, let's give one of them a try. Drag **whisperx-video-transcribe** to your workspace
7. Сonnect the **video\_url** output of your **User-Defined Input Form** node to the **Url** input of your **whisperx-video-transcribe** node.
8. Execute the **whisperx-video-transcribe** node. It will take a while to transcribe a video; the longer the video is, the more time is needed. \
9. We have a transcription, let's summarize it. Start typing `chatgpt processor` in the left panel and drag a **ChatGPT Processor** node to your workspace.\
10. Go to settings of the **ChatGPT Processor** node. \
\
You can choose ChatGPT versions with the **Model** dropdown. Let's choose gpt-4 because the transcript could be too long for older versions
11. Now it's time to prepare ChatGPT for receiving instructions. \
\
In the node settings, go to **Messages** section, click on a pencil and a new message
\
\
Change the type of the message to `System` as recommended by OpenAI to emphasize that it's a high-level instruction and to guide the model's behaviour.
12. In the **Message** field write what you need **ChatGPT** to do: something like `summarize a following video transcript`
13. Add another message (there's no need to change the message type this time). \
\
Here's a tricky part: due to the syntax of the **ChatGPT Processor** node we will need to use **Expression editor**. \
\
Click on the **#** symbol in the lower **Message** field.
\
\
You'll get the **Expression editor**. There is a list of the nodes on the left. Click on the **whisperx-video-transcribe** to see its output.
\
\
Drag the **success** output to the expression field
\
\
The result should look like this
\
\
Click the **Save** button in the **Expression editor** and don't forget to click top-right **Save** as well before leaving the settings.
14. Connect the output of the **whisperx-video-transcribe** to the input of the **ChatGPT Processor** node and run the latter by clicking the **Execute** button on the node. \
\
An important note: don't click the **Start** button on the top panel yet! If you do this, all nodes will start over, and we don't want this because video transcription is a quite time-consuming task.
15. When the **ChatGPT** finishes its work, you'll see that the text in this node is way shorter than the initial transcript. You can read it right away or copy it — just hover over the text to see the respective icon.
\
\
Of course, you can copy the full transcript as well.
16. Like storing all input data in a single node, collecting all output data in a single node is a good practice as well. For example, it will simplify matters greatly if you are going to launch the workflow via API in the future.\
\
Let's grab another **User-Defined Input Form** and add a couple of fields. Name them, for example, `full_transcript` and `summary`. Refer to paragraphs 2, 3 and 4 of this guide if needed.\
17. Connect the output of the **whisperx-video-transcribe** node to the **Full\_transcript** input and the **Success** output of the **ChatGPT Processor** to the **Summary** input respectively. Run the final node by clicking the **Execute** button\
18. Now both a full transcript and a summary are stored in one node.
And finally: once you have the workflow built, there's no need to run nodes one by one anymore. To transcribe and summarize another video, simply change the link in the **Video\_url** field of the **User-Defined Input Form** and use the **Start** button on the top bar this time.
