79
8
DeepSeek V4 Flash 0731: 82.7% on Terminal-Bench 2.1 with a public harness (antigma.ai)
3
Claude Code Cut Their System Prompt by 80%. Does That Work for Small Models Too? (antigma.ai)
8
Open-weight 27B hits 38% on Terminal-Bench 2.0 (Opus 4.1 hit 38% in Aug 2025) (antigma.ai)
2