bubble boi
@bubbleboi
mascot @thru_xyz. no investment advice.
Joined June 2017
6.3K Following 111.5K Followers
Oh my god
On Terminal-Bench 4.0, which evaluates coding agents on realistic, resource-constrained terminal tasks using strict task-specific tests, K3.1 Pro scores 65.7%, ahead of GPT-6 Astra (58.2%) and narrowly behind Claude Opus 5.5 (66.4%).
Show more
