Benchmark Almanac

AgReason

Gemini 2.5 Flash Sep 2025 36% leads since September 2025
Not in model cards

Models answer 100 farm management questions using local growing conditions. A judge compares individual claims with expert answers. A question passes at statement-level F1 of 0.80.

The authors publish 23 scored configurations, including GPT-4o, Gemini 2.5 Flash, Grok-3 Beta, Llama 4 Maverick and QwQ-32B.

Released
May 2025
Kind
Science questions
Level
University
Scored
Graded by a model
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Written by experts
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Gemini 2.5 Flash Sep 2025 #112 on the capability index
Last model added
September 2025
Scores refreshed
November 20th, 2026
23
Entries on its own leaderboard, 11 models we track

Gemini 2.5 Flash leads with 36% among models released by September 2025 Gemini 2.5 Flash leads with 36% among models released by September 2025

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in September 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Gemini 2.5 Flash Sep 2025 36%
  2. 2 Grok 3 22%
  3. 3 o1 20%
  4. 4 Phi-3-14b (SFT) 13%
  5. 5 Qwen-2.5-14b (SFT) 8.00%
  6. 6 Mistral 7B (Full SFT) 7.00%
  7. 7 DeepSeek-V3 Mar 2025 7.00%
  8. 8 Phi-3-8b (SFT) 6.00%
  9. 9 QwQ-32B 5.00%
  10. 10 GPT-4o Mar 2025 5.00%
  11. 11 LLaMA-3 8B (Full SFT) 4.00%
  12. 12 Llama 4 Maverick 4.00%
  13. 13 LLaMA-3 8B (SFT) 3.00%
  14. 14 Mistral 7B (SFT) 3.00%
  15. 15 Qwen-2.5-8b (SFT) 3.00%
  16. 16 Qwen2.5-14B 3.00%
  17. 17 Grok-2 Dec 2024 3.00%
  18. 18 Phi-3-14b 1.00%
  19. 19 phi3-8b 1.00%
  20. 20 Llama 4 Scout 1.00%
  21. 21 Mistral 24B 0%
  22. 22 Qwen-2.5-8b 0%
  23. 23 Mistral 7B v0.1 0%

Source: Official AgReason leaderboard

#EntryScoreSettingDateSource
1 Gemini 2.5 Flash Sep 2025 36% Sep 2025 Official leaderboard
2 Grok 3 22% Apr 2025 Official leaderboard
3 o1 20% Dec 2024 Official leaderboard
4 Phi-3-14b (SFT) 13% Official leaderboard
5 Qwen-2.5-14b (SFT) 8.00% Official leaderboard
6 Mistral 7B (Full SFT) 7.00% Official leaderboard
7 DeepSeek-V3 Mar 2025 7.00% Mar 2025 Official leaderboard
8 Phi-3-8b (SFT) 6.00% Official leaderboard
9 QwQ-32B 5.00% Mar 2025 Official leaderboard
10 GPT-4o Mar 2025 5.00% Mar 2025 Official leaderboard
11 LLaMA-3 8B (Full SFT) 4.00% Official leaderboard
12 Llama 4 Maverick 4.00% Apr 2025 Official leaderboard
13 LLaMA-3 8B (SFT) 3.00% Official leaderboard
14 Mistral 7B (SFT) 3.00% Official leaderboard
15 Qwen-2.5-8b (SFT) 3.00% Official leaderboard
16 Qwen2.5-14B 3.00% Sep 2024 Official leaderboard
17 Grok-2 Dec 2024 3.00% Dec 2024 Official leaderboard
18 Phi-3-14b 1.00% Official leaderboard
19 phi3-8b 1.00% Official leaderboard
20 Llama 4 Scout 1.00% Apr 2025 Official leaderboard
21 Mistral 24B 0% Official leaderboard
22 Qwen-2.5-8b 0% Official leaderboard
23 Mistral 7B v0.1 0% Sep 2023 Official leaderboard
#ModelScoreSettingDateSource
1 Gemini 2.5 Flash Sep 2025 36% Sep 2025 Official leaderboard
2 Grok 3 22% Apr 2025 Official leaderboard
3 o1 20% Dec 2024 Official leaderboard
4 DeepSeek-V3 Mar 2025 7.00% Mar 2025 Official leaderboard
5 QwQ-32B 5.00% Mar 2025 Official leaderboard
6 GPT-4o Mar 2025 5.00% Mar 2025 Official leaderboard
7 Llama 4 Maverick 4.00% Apr 2025 Official leaderboard
8 Qwen2.5-14B 3.00% Sep 2024 Official leaderboard
9 Grok-2 Dec 2024 3.00% Dec 2024 Official leaderboard
10 Llama 4 Scout 1.00% Apr 2025 Official leaderboard
11 Mistral 7B v0.1 0% Sep 2023 Official leaderboard

Rankings by part

We read 10 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Abiotic Harvest Qs, Plant & Seed Health Qs and Abiotic Soil Qs. Gemini 2.5 Flash leads 7 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

Abiotic Harvest Qs

11 models we track have a score here.

  1. 1 o1 ↑2 70%
  2. 2 Grok 3 60%
  3. 3 Gemini 2.5 Flash Sep 2025 ↓2 40%
  4. 4 DeepSeek-V3 Mar 2025 20%
  5. 5 Qwen2.5-14B ↑3 10%
  6. 6 GPT-4o Mar 2025 10%
  7. 7 Mistral 7B v0.1 ↑4 0%
  8. 8 Grok-2 Dec 2024 ↑1 0%
  9. 9 QwQ-32B ↓4 0%
  10. 10 Llama 4 Maverick ↓3 0%

2 models have held number one on the official leaderboard since its release in May 2025

Lab claims count from the day the card came out, independent runs from the model’s release.

xAI Google DeepMind
Own leaderboard Not tested since Sep 2025

Source: Official AgReason leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Gemini 2.5 Flash Grok 3 Qwen2.5-14B Mistral 7B v0.1 0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Gemini 2.5 Flash Grok 3

Source: Official AgReason leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Science and reasoning

All 203 in Science and reasoning
  • GPQA Diamond In use 198 graduate-level 'Google-proof' multiple-choice questions in biology, chemistry and physics. Gemini 4 Carbon 97%
  • Terminal-Bench-Science In use 70 scientific research-workflow tasks across five science areas, run in Harbor containers. Hidden tests give the resolution rate. GPT-6 Astra 68.1%
  • BioMysteryBench In use Bioinformatics problems on anonymized real datasets that require real analysis with tools. Rubric-graded, with human-solvable and human-difficult splits. Claude Fable 5.2 83.7%
  • FrontierScience In use Models solve physics, chemistry, and biology problems in an Olympiad track graded by short answer and a Research track graded by 10-point rubrics. Intern-S2-397B 87%
  • HiPhO In use Models solve theory problems from 2024 and 2025 physics olympiad exams, some with diagrams. Official marking schemes grade answers and steps, mapped to medal levels. Gemini 3 Pro 78.3%
  • LABBench2 In use 11 biology research sub-tasks (literature, figures, tables, patents, trials and more) with tool access. Macro-averaged accuracy. GPT-5.2 60.9%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 14 months ago, in September 2025, for Gemini 2.5 Flash. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (23 entries).
  • Pass Models scored 11 models we track, plus 12 other entries on its leaderboard.
  • Warning Newest score The newest came 14 months ago, in September 2025, for Gemini 2.5 Flash.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 2.5 Flash (#112).
  • Pass Headroom The best is 36%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on baskargroup.github.io: 11 rows for models we track and 12 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check