Customize the Recursive Character Text Splitter Node

[Richard_Me]

1. Is there a way to send a custom parameter to OpenAI Embeddings in n8n for specifying a custom dimension (e.g., 256)?
   With the new text-embedding-3-large model from OpenAI, there’s an option to set a custom dimensional parameter (like 256). How can I configure this in n8n’s OpenAI node, or is there a workaround using HTTP Request nodes to achieve this?

2. How can I ensure that the Recursive Character Text Splitter in n8n keeps the section title in every chunk?
   When using the Recursive Character Text Splitter in the default data loader, I would like each chunk to retain the section title for better context in downstream processing. Is there a setting to achieve this, or would I need a custom solution, like adding a Code node?

[ria]

Hi @Richard_Me

setting custom dimensions is actually something we have just been working on! Should be released soon

feat(Embeddings OpenAI Node): Add dimensions option

to answer your 2nd question, this is unfortunately not possible at the moment but definitely something we will be working on!