Token cost is counted even if langchains LLM cache is utilized

[arthurGrigo]

Not sure if this behaviour is intended as it is but when I use langchains LLM cache and get_openai_callback() it shows me total cost of 0$ for prompts I executed before.

In langfuse I see the cost of the prompt as it would have been sent so the API.

@marcklingen Bug or feature?

[marcklingen]

I’d go for bug here, thanks for reporting!

Desiree behavior

• Token counts are not 0
• Cost is 0

Tracking this here: #2112