Vanish TPS
Tokens per second, in the HUD.
Shows tokens per second for the AI's last response. Handy for benchmarking models and spotting slow providers.
Features
- 01Live TPS while a response streams (e.g., ~12.4 t/s).
- 02Final tally locks in once the response finishes, including total token count.
- 03Read-only HUD entry — no toggles, no fuss.