Turn long audio into ready content in minutes. The flow transcribes an MP3, writes a short summary, translates it to German, and drafts a cover image idea with matching visuals. It suits podcast teams, marketers, and social managers who need fast turnaround.

After you click run, a file reader loads an MP3 and Whisper turns speech into text. That text moves through ChatGPT nodes to produce a clear tl dr and a German version. A code node builds a message array and an HTTP Request node calls the chat API directly for more control. Another branch asks ChatGPT to write an image prompt and then DALLE 2 returns four cover images. A set node and a chat node shape structured output, and an HTML node builds a clean block with inline SVG for web or email.

You only need an OpenAI account and API key. Add credentials in n8n and test one branch at a time to avoid long runs. Teams can cut content prep from hours to minutes, reuse the HTML block on sites, and ship multilingual posts for each episode or webinar. It is a good fit for podcast recaps, webinar highlights, and short news briefings.

• Manual start so you control when each branch runs and avoid extra API calls.
• Read MP3 and transcribe speech to text using Whisper for quick, accurate text.
• Create a short tl dr with ChatGPT to capture key points fast.
• Translate the summary to German with a separate ChatGPT step.
• Build a custom message array in a code node and call the chat API via HTTP Request for fine control.
• Generate an image prompt with ChatGPT and return four cover images from DALLE 2.
• Assemble an HTML block with inline SVG for a ready to paste section on web or email.
• Include legacy text completion and edit examples to support teams moving from older models.

• Reduce manual work from 2 hours per episode to 15 minutes
• Automate transcription, summary, translation, and images in one run
• Keep tone and structure consistent across posts and languages
• Produce web ready HTML you can paste into sites and emails
• Scale to handle 10 recordings per day with the same setup

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with OpenAI. See the Tools Required section above for links to create accounts with these services.
3. In n8n Cloud, open Credentials and create a new OpenAI credential using your API key from the OpenAI dashboard. Give it a clear name such as OpenAI Main.
4. Open each OpenAI node and select your OpenAI credential in the Credential to connect with field. Do this for Whisper, ChatGPT, and the DALLE 2 image node.
5. Open the HTTP Request node that calls the Chat Completions API. In the Authentication or Credentials section, choose the same OpenAI credential. If a credential is not available, add an Authorization header with Bearer followed by your API key and set Content Type to application/json.
6. Double click the LoadMP3 node and set the file path to your MP3. Use a small test file first to validate speed and cost.
7. If you are using real audio, disable the Text example node so the transcribed text flows into the summary steps.
8. Run the workflow manually and execute only the next node in one branch to test transcription. Confirm you see text in the output.
9. Execute the TLDR node and then the translation node. Check that the German text appears as a separate field.
10. Run the image prompt node and then the image generation node. Verify that four images are produced as binary outputs.
11. Test the HTML node to view the generated HTML and inline SVG. Copy the output into a test page or email builder to confirm rendering.
12. If runs are slow or you see errors, disable branches you do not need. Reduce max tokens in Chat nodes and try again. Check your OpenAI usage limits if requests fail.