Benchmark Almanac

Can AI close the books? FP&A reasoning

Gemini 2.5 Pro Jun 2025 100% leads since October 2026
Not in model cards

Models complete synthetic financial close tasks, including accruals and reconciliations. The grader checks calculations and whether reviewer claims agree with the books.

Kaggle publishes scores for 40 distinct model versions.

Made by
Marco Polo
Released
October 2026
Kind
Finance
Level
University
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Generated
Refresh
One fixed set
Run by
Anyone
In views
Law and finance
Strongest model tested
Claude Opus 5.5 #5 on the capability index
Last model added
October 2026
Scores refreshed
November 20th, 2026
40
Entries on its own leaderboard, 38 models we track

18 models share the top score of 100% 18 models share the top score of 100%

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.

  1. 1 Gemini 2.5 Pro Jun 2025 100%
  2. 2 Gemini 3 Flash 100%
  3. 3 GLM-5 100%
  4. 4 Gemini 3.1 Pro 100%
  5. 5 Gemma 4 26B A4B 100%
  6. 6 Gemma 4 31B IT 100%
  7. 7 GPT-5.5 100%
  8. 8 Claude Sonnet 5 100%
  9. 9 GPT-5.6 Luna 100%
  10. 10 GPT-5.6 Sol 100%
  11. 11 Gemini 3.6 Flash 100%
  12. 12 Claude Opus 5 100%
  13. 13 Gemini 3.8 Flash 100%
  14. 14 Claude Opus 5.5 100%
  15. 15 GPT-6 Luna 100%
  16. 16 GPT-6 Sol 100%
  17. 17 GPT-6.1 Sol 100%
  18. 18 Claude Haiku 5.5 100%
  19. 19 gpt-oss-120b 96.4%
  20. 20 Gemini 2.5 Flash Sep 2025 92.9%
  21. 21 gpt-oss-20b 85.7%
  22. 22 Gemini 3.1 Flash-Lite 75%
  23. 23 GPT-5.4 75%
  24. 24 Claude Opus 4.5 67.9%
  25. 25 Claude Opus 4.7 67.9%
  26. 26 Gemini 3.5 Flash-Lite 67.9%
  27. 27 GPT-5.4 Mini 53.6%
  28. 28 Grok 4.20 non-reasoning 50%
  29. 29 Qwen3-Coder-480B-A35B 46.4%
  30. 30 Qwen3-Next-80B-A3B 42.9%
  31. 31 Claude Sonnet 5.5 42.9%
  32. 32 Claude Haiku 4.5 39.3%
  33. 33 Qwen 3 235B A22B Instruct 2506 39.3%
  34. 34 Qwen 3 Next 80B Instruct 39.3%
  35. 35 Claude Opus 4.8 35.7%
  36. 36 GPT-5.4 Nano 32.1%
  37. 37 Claude Sonnet 4.6 21.4%
  38. 38 Claude Opus 4.6 14.3%
  39. 39 DeepSeek-R1 May 2025 0%
  40. 40 Gemini 3.5 Flash 0%

Source: Official Can AI close the books? FP&A reasoning leaderboard

Source: Official Can AI close the books? FP&A reasoning leaderboard

#EntryScoreSettingReleasedRun onSource
1 Gemini 2.5 Pro Jun 2025 100% Jun 2025 Oct 2026 Official leaderboard
2 Gemini 3 Flash 100% Dec 2025 Oct 2026 Official leaderboard
3 GLM-5 100% Feb 2026 Oct 2026 Official leaderboard
4 Gemini 3.1 Pro 100% Feb 2026 Oct 2026 Official leaderboard
5 Gemma 4 26B A4B 100% Apr 2026 Oct 2026 Official leaderboard
6 Gemma 4 31B IT 100% Apr 2026 Oct 2026 Official leaderboard
7 GPT-5.5 100% Apr 2026 Oct 2026 Official leaderboard
8 Claude Sonnet 5 100% Jun 2026 Oct 2026 Official leaderboard
9 GPT-5.6 Luna 100% Jul 2026 Oct 2026 Official leaderboard
10 GPT-5.6 Sol 100% Jul 2026 Oct 2026 Official leaderboard
11 Gemini 3.6 Flash 100% Jul 2026 Oct 2026 Official leaderboard
12 Claude Opus 5 100% Jul 2026 Oct 2026 Official leaderboard
13 Gemini 3.8 Flash 100% Sep 2026 Oct 2026 Official leaderboard
14 Claude Opus 5.5 100% Sep 2026 Oct 2026 Official leaderboard
15 GPT-6 Luna 100% Sep 2026 Oct 2026 Official leaderboard
16 GPT-6 Sol 100% Sep 2026 Oct 2026 Official leaderboard
17 GPT-6.1 Sol 100% Sep 2026 Oct 2026 Official leaderboard
18 Claude Haiku 5.5 100% Oct 2026 Oct 2026 Official leaderboard
19 gpt-oss-120b 96.4% Aug 2025 Oct 2026 Official leaderboard
20 Gemini 2.5 Flash Sep 2025 92.9% Sep 2025 Oct 2026 Official leaderboard
21 gpt-oss-20b 85.7% Aug 2025 Oct 2026 Official leaderboard
22 Gemini 3.1 Flash-Lite 75% Mar 2026 Oct 2026 Official leaderboard
23 GPT-5.4 75% Mar 2026 Oct 2026 Official leaderboard
24 Claude Opus 4.5 67.9% Nov 2025 Oct 2026 Official leaderboard
25 Claude Opus 4.7 67.9% Apr 2026 Oct 2026 Official leaderboard
26 Gemini 3.5 Flash-Lite 67.9% Jul 2026 Oct 2026 Official leaderboard
27 GPT-5.4 Mini 53.6% Mar 2026 Oct 2026 Official leaderboard
28 Grok 4.20 50% non-reasoning Feb 2026 Oct 2026 Official leaderboard
29 Qwen3-Coder-480B-A35B 46.4% Jul 2025 Oct 2026 Official leaderboard
30 Qwen3-Next-80B-A3B 42.9% Sep 2025 Oct 2026 Official leaderboard
31 Claude Sonnet 5.5 42.9% Sep 2026 Oct 2026 Official leaderboard
32 Claude Haiku 4.5 39.3% Oct 2025 Oct 2026 Official leaderboard
33 Qwen 3 235B A22B Instruct 2506 39.3% Oct 2026 Official leaderboard
34 Qwen 3 Next 80B Instruct 39.3% Oct 2026 Official leaderboard
35 Claude Opus 4.8 35.7% May 2026 Oct 2026 Official leaderboard
36 GPT-5.4 Nano 32.1% Mar 2026 Oct 2026 Official leaderboard
37 Claude Sonnet 4.6 21.4% Feb 2026 Oct 2026 Official leaderboard
38 Claude Opus 4.6 14.3% Feb 2026 Oct 2026 Official leaderboard
39 DeepSeek-R1 May 2025 0% May 2025 Oct 2026 Official leaderboard
40 Gemini 3.5 Flash 0% May 2026 Oct 2026 Official leaderboard
#ModelScoreSettingReleasedRun onSource
1 Gemini 2.5 Pro Jun 2025 100% Jun 2025 Oct 2026 Official leaderboard
2 Gemini 3 Flash 100% Dec 2025 Oct 2026 Official leaderboard
3 GLM-5 100% Feb 2026 Oct 2026 Official leaderboard
4 Gemini 3.1 Pro 100% Feb 2026 Oct 2026 Official leaderboard
5 Gemma 4 26B A4B 100% Apr 2026 Oct 2026 Official leaderboard
6 Gemma 4 31B IT 100% Apr 2026 Oct 2026 Official leaderboard
7 GPT-5.5 100% Apr 2026 Oct 2026 Official leaderboard
8 Claude Sonnet 5 100% Jun 2026 Oct 2026 Official leaderboard
9 GPT-5.6 Luna 100% Jul 2026 Oct 2026 Official leaderboard
10 GPT-5.6 Sol 100% Jul 2026 Oct 2026 Official leaderboard
11 Gemini 3.6 Flash 100% Jul 2026 Oct 2026 Official leaderboard
12 Claude Opus 5 100% Jul 2026 Oct 2026 Official leaderboard
13 Gemini 3.8 Flash 100% Sep 2026 Oct 2026 Official leaderboard
14 Claude Opus 5.5 100% Sep 2026 Oct 2026 Official leaderboard
15 GPT-6 Luna 100% Sep 2026 Oct 2026 Official leaderboard
16 GPT-6 Sol 100% Sep 2026 Oct 2026 Official leaderboard
17 GPT-6.1 Sol 100% Sep 2026 Oct 2026 Official leaderboard
18 Claude Haiku 5.5 100% Oct 2026 Oct 2026 Official leaderboard
19 gpt-oss-120b 96.4% Aug 2025 Oct 2026 Official leaderboard
20 Gemini 2.5 Flash Sep 2025 92.9% Sep 2025 Oct 2026 Official leaderboard
21 gpt-oss-20b 85.7% Aug 2025 Oct 2026 Official leaderboard
22 Gemini 3.1 Flash-Lite 75% Mar 2026 Oct 2026 Official leaderboard
23 GPT-5.4 75% Mar 2026 Oct 2026 Official leaderboard
24 Claude Opus 4.5 67.9% Nov 2025 Oct 2026 Official leaderboard
25 Claude Opus 4.7 67.9% Apr 2026 Oct 2026 Official leaderboard
26 Gemini 3.5 Flash-Lite 67.9% Jul 2026 Oct 2026 Official leaderboard
27 GPT-5.4 Mini 53.6% Mar 2026 Oct 2026 Official leaderboard
28 Grok 4.20 50% non-reasoning Feb 2026 Oct 2026 Official leaderboard
29 Qwen3-Coder-480B-A35B 46.4% Jul 2025 Oct 2026 Official leaderboard
30 Qwen3-Next-80B-A3B 42.9% Sep 2025 Oct 2026 Official leaderboard
31 Claude Sonnet 5.5 42.9% Sep 2026 Oct 2026 Official leaderboard
32 Claude Haiku 4.5 39.3% Oct 2025 Oct 2026 Official leaderboard
33 Claude Opus 4.8 35.7% May 2026 Oct 2026 Official leaderboard
34 GPT-5.4 Nano 32.1% Mar 2026 Oct 2026 Official leaderboard
35 Claude Sonnet 4.6 21.4% Feb 2026 Oct 2026 Official leaderboard
36 Claude Opus 4.6 14.3% Feb 2026 Oct 2026 Official leaderboard
37 DeepSeek-R1 May 2025 0% May 2025 Oct 2026 Official leaderboard
38 Gemini 3.5 Flash 0% May 2026 Oct 2026 Official leaderboard

Number one over time

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

Google DeepMind
Own leaderboard Passed 85% in Oct 2026, by Gemini 2.5 Pro

Source: Official Can AI close the books? FP&A reasoning leaderboard

Gemini 2.5 Pro on the official leaderboard passed 85% in June 2025

Models released before it came out already scored that high. 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% 2025 2026 Near the ceiling (85%) Gemini 2.5 Pro DeepSeek-R1 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) Gemini 2.5 Pro DeepSeek-R1

Source: Official Can AI close the books? FP&A reasoning 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 Law, finance and medicine

All 637 in Law, finance and medicine
  • MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy. Qwen 3.8 Max 80.4%
  • HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
  • HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
  • FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy. Ling-3.1-flash 81.7%
  • KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. o1-preview 92.7%
  • Harvey LAB-AA v1.1 In use Artificial Analysis's run of Harvey's Legal Agent Benchmark on Harvey's newer private set of 120 tasks across 24 practice areas. A deliverable with a material hallucination scores zero, and the headline is the share of judges who find every rubric criterion passed. GPT-6.2 Astra 15%

Data audit

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (40 entries).
  • Pass Models scored 38 models we track, plus 2 other entries on its leaderboard.
  • Pass Newest score The newest is from October 2026, for Claude Haiku 5.5.
  • Pass Frontier models 3 of the top 10 have a score, led by Claude Opus 5.5 (#5).
  • Note Headroom Saturated. The best is 100%, past 85% since June 2025.
Leaderboard import 3 pass
  • Pass Importer We read the board on kaggle.com: 38 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