LongDS v1.1 Full - Geoscience: leaderboard

Metric: Accuracy (%; judged turn accuracy macro-averaged over 19 tasks), each model under one agentic coding harness. Source: zjunlp.github.io. Saturation forecast: Around March 2028. 6 models tracked.

Top models

#ModelScore
1GPT-5.6 Sol56.18
2Kimi K346.02
3Qwen 3.8 Max (0902)43.85
4GLM-5.242.51
5DeepSeek V4 Pro34.3

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