Benchmark Almanac

FieldWorkArena V2.0

GPT-5.2 52% leads since December 2025
Not in model cards

Agents inspect workplace images, videos, manuals, and logs to plan actions and report operational events. The evaluator compares their answers and actions with reference results. It combines correctness judgments with partial credit for numerical agreement.

Version 2.0 adds retail tasks to the factory and warehouse releases. Version 2.1 only clarifies query wording. The maker requires a data-access application, and its public leaderboard is still listed as forthcoming. The paper reports the earlier factory and warehouse evaluation.

Released
June 2025
Kind
Tasks from real jobs
Level
University
Scored
Graded by a model
Answers
Agent task
Measures
Capability
Language
English
In and out
Text, Images, Video, Documents → Text, Actions
Interaction
Agentic
Environment
APIs and tools
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Agents, Vision
Strongest model tested
GPT-5.2 #51 on the capability index
Last model added
December 2025
Scores refreshed
November 20th, 2026

In the authors’ words

This paper introduces FieldWorkArena, a benchmark for agentic AI targeting real-world field work. With the recent increase in demand for agentic AI, they are built to detect and document safety hazards, procedural violations, and other critical incidents across real-world manufacturing and retail environments. Whereas most agentic AI benchmarks focus on performance in simulated or digital environments, our work addresses the fundamental challenge of evaluating agents in the real-world. In this paper, we improve the evaluation function from previous methods to assess the performance of agentic AI in diverse real-world tasks. Our dataset comprises on-site captured images/videos in factories, warehouses and retails. Tasks were meticulously developed through interviews with site workers and managers. Evaluation results confirmed that performance evaluation considering the characteristics of Multimodal LLM (MLLM) such as GPT-4o is feasible. Furthermore, this study identifies both the effectiveness and limitations of the proposed new evaluation methodology. The complete dataset and evaluation program are publicly accessible on the website (https://en-documents.research.global.fujitsu.com/fieldworkarena/)

Abstract of FieldWorkArena: Agentic AI Benchmark for Real Field Work Tasks (arXiv:2505.19662, CC0)
5
Entries in its paper, 3 models we track

GPT-5.2 leads with 52% among models released by December 2025 GPT-5.2 leads with 52% among models released by December 2025

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. 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 December 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 GPT-5.2 52%
  2. 2 Gemini-2.5 (Pro) (snapshot unspecified) 46%
  3. 3 GPT-5.1 43%
  4. 4 Gemini-2.5 (Flash) (snapshot unspecified) 40%
  5. 5 GPT-4o Aug 2024 35%

Source: The FieldWorkArena V2.0 paper

Source: The FieldWorkArena V2.0 paper

#EntryScoreSettingDateSource
1 GPT-5.2 52% Dec 2025 Official leaderboard
2 Gemini-2.5 (Pro) (snapshot unspecified) 46% Official leaderboard
3 GPT-5.1 43% Nov 2025 Official leaderboard
4 Gemini-2.5 (Flash) (snapshot unspecified) 40% Official leaderboard
5 GPT-4o Aug 2024 35% Aug 2024 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-5.2 52% Dec 2025 Official leaderboard
2 GPT-5.1 43% Nov 2025 Official leaderboard
3 GPT-4o Aug 2024 35% Aug 2024 Official leaderboard

Rankings by part

We read 9 parts from the benchmark’s paper, each ranking models on one part of the test, such as Perception, Decision making and Combination. GPT-5.2 leads 8 of them, and GPT-5.1 leads the other one. Arrows show how far a model moved from its overall rank among the models we track.

Perception

3 models we track have a score here.

  1. 1 GPT-5.2 52%
  2. 2 GPT-5.1 45%
  3. 3 GPT-4o Aug 2024 35%

3 models have held number one in the benchmark’s paper since its release in June 2025

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

OpenAI
Benchmark paper Not tested since Dec 2025
GPT-5.2 52%

Source: The FieldWorkArena V2.0 paper

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
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-5.2 GPT-5.1 GPT-4o 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-5.2 GPT-5.1

Source: Official FieldWorkArena V2.0 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 Economic work

All 210 in Economic work
  • JobBench In use 130 tasks from 35 occupations workers most want delegated. Agent works through messy reference files, graded by fact-anchored binary rubrics. GLM-5.3 61.4%
  • GDPval In use Real work tasks from 44 occupations with reference files. Experts blindly compare model deliverables with expert work. Metric is win-or-tie rate. Seed2.1 Pro 87.9%
  • GDPval-AA v2.1 In use Agents with shell and web access complete 220 GDPval gold tasks from 44 occupations. Blind pairwise comparisons give Elo ratings. Claude Opus 5.5 1,846
  • AA-Briefcase v1.1 In use Version 1.1 of Artificial Analysis's multi-hour office work evaluation, with Elo ratings fitted by Crowd-BT. Claude Opus 5.5 1,822
  • Agents' Last Exam v1.1 In use Agents complete 122 public long-horizon professional workflow tasks across 55 sub-industries. Deterministic graders check outcomes against hidden references. GPT-6.2 Astra 47.5%
  • CoWorkBench In use Agent completes long-horizon productivity tasks across computer science, finance, law and medicine. Qwen in-house scoring (method unpublished). Claude Fable 5 75.9%

Data audit

8 of 10 checks pass, 2 warnings. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (5 entries).
  • Pass Models scored 3 models we track, plus 2 other entries on its leaderboard.
  • Warning Newest score The newest came 11 months ago, in December 2025, for GPT-5.2.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.2 (#51).
  • Pass Headroom The best is 52%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 3 rows for models we track and 2 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