Skip to content

UCSF (and maybe external) token usage mismatch #1

@paulsonak

Description

@paulsonak

Please explain the motivation behind the feature request.
On the Versa API tableau dashboard, I have a usage one day of 66M tokens from aws-sdk-rust/1.3.11 which i believe is BioRouter. However, when I looked at my chats from that day, it showed many fewer tokens:
Image

I created a diagnostics bundle from my highest usage chat and found that there are total tokens, and accumulated total tokens:

  "total_tokens": 218562,
  "input_tokens": 217885,
  "output_tokens": 677,
  "accumulated_total_tokens": 13355163,
  "accumulated_input_tokens": 13292340,
  "accumulated_output_tokens": 62823,

Describe the solution you'd like
I think accumulated_total_tokens per conversation is more informative than total_tokens. Even better would be tokens - per - model (sometimes I switch models in the same thread - still testing and settling between GPT and Claude) or being able to review/query this data in a light database. Also, since I can pick up conversations over multiple days, a per-day or so-far-this-month report of usage would be nice to keep an eye on where I'm at vs the UCSF subsidized limit.

Describe alternatives you've considered
I can pull all the diagnostics myself, and sort by day, etc, but it looks like you're already attempting some of this with your Pricing Data feature so hopefully this would be a useful addition that isn't too onerous to implement.

Additional context
Versa API cost dashboard screenshot:
Image

  • I have verified this does not duplicate an existing feature request

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions