How to use a big file as input parameter?

[SamYuan1990]

Pre-submission Checklist

• I have checked that this question would not be more appropriate as an issue in a specific repository
• I have searched existing discussions and documentation for answers

Question Category

• Protocol Specification
• SDK Usage
• Server Implementation
• General Implementation
• Documentation
• Other

Your Question

hello here,

My question as I designed a MCP server, which used for i18n, I used my prompts to translate hugging face document in previous, and here is link https://github.com/SamYuan1990/i18n-agent-action

During make it as an agent, I noticed that I can support further more scope, not only technological docs but also with sound to text support and text to sound impls “simultaneous interpretation” like Apple's new ipod in recent.

For text, I can make it simple as input parameter is text.
For audio/voice, my original plan is to make it in base64 through network connection.
But here is the problem....

either big text(one of my friend try with my agent with a digital book about 10k words)..., or a voice file(about 200~300k) after base64 will out of LLM's content window.

So how MCP Protocol support this case?

[hasani114]

I haven't been able to find an answer for this either. It has been fine for local servers (you can just provide the path to the file) but for remote servers it seems like you would need to upload the files to a server and then insert the URL for that.