User level reporting
planned
S
Stephen
Having the ability to report on a per user level of usage, model selection and credit usage over a specific period.
I often have clients asking for reports but lack the ability to dig deeper. I have setup some custom reports using the API calls, but the user level endpoint doesn't have the granular details I'd prefer.
being able to identify the users defaulting to GPT5.5 or Opus 4.7,4.8 who use 50% of the tenants credits is becoming a pressing matter.
M
Mark Leeper
Include client admins in this
M
Mark Leeper
Merged in a post:
Reporting on a per user basis instead of just seeing who the top user is.
D
Donald McArthur
M
Mark Leeper
Merged in a post:
Need Internal Admin Visibility!
J
Jennifer Albrecht
I am the internal IT Manager and I have no visibility into our account. I need to be able to see usage per user, per LLM, etc... I need to see what our MSP sees (to a point) in order to fully be able to manage our users/usage. this is very frustrating.
M
Mark Leeper
Merged in a post:
I love the new changes, however, with the restructure of the LLM's, it doesn't show the token usage per LLM anymore. We really try to push our users to use the lower cost models when appropriate to save on token usage. Now there is no way to tell which ones are heavy toke users. thanks.
J
Jennifer Albrecht
J
Julia Carroll
updated the status to
planned
N
Noah Jones
This is a request we have had from multiple clients. With custom roles you can provision access to the usage dashboard, however they can only see power users and general usage metrics, not a user level breakdown necessary to support their internal processes. The only way to give them visibility into user level credit usage is to assign the permission to manage users, which is not least privilege access for what they need.
E
Edward Macmillan
This is gaining more traction with our clients. Having a reporting mechanism to let each individual user see their credit consumption is more important now than ever. Also, we need to be able to see more than a rolling 30 days of daily usage. I would like to recommend going back at least 90 days as a suggestion.
D
Dan Davis
One client just asked for this before signing up. If she can't figure out which employee is using all of the shared credits, she doesn't want to use the service, full stop. I agree we need better granularity in the reporting. The data is absolutely there already, if you can report on Power Users!
S
Stephen
Especially as the latest models are super credit hungry, the Interns can now burn 12x credits with Fable 5.
N
Nicole Lefsky
With respect to model selection, even if we enable the auto LLM selector by default, users may choose to use frontier models or those that use more tokens then they actually need for the task. We can't guide them if we can't see what models they're using.
Load More
→