TerraLogic - Spatial Relations Toolkit: leaderboard

Metric: End-to-end score (%) of tool-augmented predictions on the Spatial Relations toolkit over the 545 TerraLogic remote-sensing tasks, each model driving the HieraPlan tool agent under a ReAct-style protocol in OpenCompass; higher is better. Source: arxiv.org. Saturation forecast: Around June 2027. 12 models tracked.

Top models

#ModelScore
1Qwen 2.5 7B Instruct82.42
2GPT-4o77.95
3Llama 3 70B Instruct62.54
4Qwen 2.5 32B Instruct61.47
5InternLM3-8B-Instruct60.7
6DeepSeek V357.93
7GPT-3.553.3
8Phi-3 Mini 4K Instruct51.74
9Mistral 7B Instruct (v0.2)45.23
10Gemini 2.5 Flash41.51
11Yi-1.5-6B-Chat36.02

Interactive version: theaggregate.ai/benchmark?slug=terralogic-spatial-relations-toolkit · How It Works · Data refreshed daily, snapshot 2026-09-29.