AgroTools: leaderboard

End-to-end agent mode on AgroTools (2026), image-grounded agricultural questions (pest and disease diagnosis, counting and measurement, change detection, visualization) with 14 executable agricultural tools: the model plans, calls the tools and answers. Mean slot-based final answer score from deterministic matching and a GPT-4o judge, scaled to 0-100; higher is better.

Metric: Final answer score (0-100) in end-to-end mode (the agent plans, calls the executable tools and answers with no gold history; Lagent ReAct schema, temperature 0.1) on the AgroTools shared subset (image-grounded agricultural queries from 12 public datasets, excluding raster-processing and plot-generation queries); answers are normalized into slots by a GPT-4o extractor, closed slots scored by deterministic matching and open-text slots by a GPT-4o judge against reference key points, each query scored 0-1 and averaged (times 100, so 0-100); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-5.462.74#135
2GLM-4.6V52.15#710
3Seed 2.0 Pro48.64#222
4InternVL3.5-14B43.74#1504
5Qwen 3 VL 8B Instruct41.64#1029
6Gemini 2.5 Pro40.68#276
7Qwen 3.5 9B39.19#828
8Qwen 3.5 35B A3B38.53#492
9InternVL3.5-8B36.18#1604
10Claude Sonnet 4.625.44#173
11Janus-Pro-7B1.86#6857
12InternVL3-8B1.61#1782
13LLaVA-v1.6-Vicuna-13B0.71#5527

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=agrotools · How It Works · Data refreshed daily, snapshot 2026-11-21.