LiteLLM

Every call through your proxy, owned and measured.

LiteLLM tracks spend per key, team and model. Valistry adds outcomes and savings, and the AI you pay for outside the proxy.

Read-only · Usage, bills and settings · No prompts or replies by default

Attribution

Valistry × LiteLLM

How much do we spend on AI?
The LiteLLM admin UI gives youThe calls sent through the proxy.
Valistry addsEvery bill and seat outside the proxy too.
Who owns each dollar?
The LiteLLM admin UI gives youSpend per key, team, user and tag.
Valistry addsOne owner for each dollar, across the proxy and every vendor.
What did the spend produce?
The LiteLLM admin UI gives youA log of every call, with tokens and cost.
Valistry addsThe cost of each resolved ticket or merged change.
Will we stay within budget?
The LiteLLM admin UI gives youA budget per key and team.
Valistry addsOne set of budgets across every vendor.
Where can we save?
The LiteLLM admin UI gives youRouting and fallbacks you set.
Valistry addsThe cheaper model for the job, checked on your own runs.

Question	The LiteLLM admin UI gives you	Valistry adds
How much do we spend on AI?	The calls sent through the proxy.	Every bill and seat outside the proxy too.
Who owns each dollar?	Spend per key, team, user and tag.	One owner for each dollar, across the proxy and every vendor.
What did the spend produce?	A log of every call, with tokens and cost.	The cost of each resolved ticket or merged change.
Will we stay within budget?	A budget per key and team.	One set of budgets across every vendor.
Where can we save?	Routing and fallbacks you set.	The cheaper model for the job, checked on your own runs.

Every endpoint, checked4endpoints read25 more covered another waySee every endpoint

Each AI call, except its content. Model, key, team, tokens and cost. Time to first token, prompt version and retries. Never the prompt or the reply.