LongDS v1.1 Full - Community: leaderboard

Metric: Accuracy (%; judged turn accuracy macro-averaged over 16 tasks), each model under one agentic coding harness. Source: zjunlp.github.io. Saturation forecast: Around December 2026. 6 models tracked.

Top models

#ModelScore
1GPT-5.6 Sol77.06
2Qwen 3.8 Max (0902)76.87
3Kimi K376
4GLM-5.274.47
5DeepSeek V4 Pro57.54

Interactive version: theaggregate.ai/benchmark?slug=longds-v1-1-full-community · How It Works · Data refreshed daily, snapshot 2026-09-25.