Token Counting Discrepancy with DeepSeek v3.1 Model

[linw1995]

I'm experiencing a token counting inconsistency when using the latest DeepSeek v3.1 model with Crush. The model returns a 400 error indicating that tokens are exceeding the limit, but the token statistics displayed in the sidebar show different (presumably lower) numbers.

Note: I'm not sure if this is a common problem affecting all providers or specifically limited to DeepSeek v3.1.

POST "https://api.deepseek.com/v1/chat/completions": 400 Bad Request {"message":"This model's maximum context length is 131072 tokens. However, you requested 143518 tokens (138518 in the messages, 5000 in the completion). Please reduce the length of the messages or

[meowgorithm]

Appreciate you flagging this one: on our radar.