Query on how ChatGPT parses HTML Inputs

Hi all!
I want to inquire regarding how ChatGPT parses HTML Inputs. I have a use-case where I have all my content in HTML format, and I need to pass this content and do analysis using the ChatGPT 4o model.

If ChatGPT converts HTML to text before processing it, then it’s best for me to parse it to text pre-emptively to avoid excessive cost on the input tokens.

However, if it handles HTML as it is, then it would be beneficial to incur the cost as it would understand context in a better way.

I couldn’t find any insights on this, and would like help on this.