https://promptfoo.dev/ logo
TokenUsage in openAI returns 0
# ❓|questions
a
When I send an evaluation request, the test is done correctly, but when looking into the tokenUsage for prompt and completion it gives 0 but when I use Anthrobic models, these values has numbers Do anyone knows why this happens?! https://cdn.discordapp.com/attachments/1387068500501069965/1387068500610125944/image.png?ex=685bffb5&is=685aae35&hm=f748fdf7f49d957a04af47dc566ee25f884c591f3f681c984afcc724e4dcc426&