Support prompt caching for LiteLLM

[steve-gore-snapdocs]

What specific problem does this solve?

When using LiteLLM as a provider and a model that supports prompt caching (claued 3.7), Roo does not use prompt caching. This is evident in the pricing shown in Roo and in the LiteLLM usage.

Additional context (optional)

Cline has added this support in cline/cline@801946f

Roo Code Task Links (Optional)

No response

Request checklist

• I've searched existing Issues and Discussions for duplicates
• This describes a specific problem with clear impact and context

Interested in implementing this?

• Yes, I'd like to help implement this feature

Implementation requirements

• I understand this needs approval before implementation begins

How should this be solved? (REQUIRED if contributing, optional otherwise)

No response

How will we know it works? (Acceptance Criteria - REQUIRED if contributing, optional otherwise)

No response

Technical considerations (REQUIRED if contributing, optional otherwise)

No response

Trade-offs and risks (REQUIRED if contributing, optional otherwise)

No response

[roomote]

I am working on implementing prompt caching support for LiteLLM. I will analyze the Cline implementation referenced in the issue and adapt it for Roo Code.

[mrubens]

Hi @steve-gore-snapdocs, we do have code that attempts to load the models from liteLLM and look at whether prompt caching is enabled. Do you see "Supports prompt caching" in the providers tab for the selected model?

[steve-gore-snapdocs]

I do see that:

However in LiteLLM I do not see any cache prompts for Roo:

But do see them for Cline: