On this day: Mar 24
3 records from this date, 2025 to 2026, gathered year by year — the most recent year first.
2026
This year · 2 recordsHarness design for long-running application developmentAnthropic Labs (Prithvi Rajasekaran)
3 agents — Planner, Generator, Evaluator; a single agent ran 20 minutes for $9 versus 6 hours for $200 with the full harness
Loop & Harness EngineeringOfficial
vLLM introduces Model Runner V2vLLM
The second rewrite of the execution layer since the V1 redesign. Serving competition moved from kernels to schedulers and execution cores
2025
1 year ago · 1 recordARC Prize announces ARC-AGI-2 and ARC Prize 2025ARC Prize
Made score and cost joint leaderboard metrics
Evaluation & BenchmarksOfficial