DeepSeek V4 Flash 0731: 82.7% on Terminal-Bench 2.1 with a public harness

an audit for the bench for the curious:

yeah, that bench has a lot of room to improve, but it is the best we can find now

Keeps throwing forbidden after logging-in to the dashboard and goes back to login screen. Tried 4 times.

Oh, it seems fine for me. I can login actually. Maybe network/server instability?

No idea..

From README for Ante:

"The core harness itself is developed in a private repository during the alpha"

lost all interest after that sentence, sad thing