Inkling | Terminal-Bench 2.1 — Thinking Machines internal coding harness

Thinking Machines Lab's Inkling evaluation results on Terminal-Bench 2.1 — Thinking Machines internal coding harness

Score: 63.8%

Reasoning level: effort=0.99

Source

Inkling (Registry No. 160001)

Thinking Machines Lab

Terminal-Bench 2.1 — Thinking Machines internal coding harness