Benchmark Almanac

BizFinBench v2

GPT-5 61.5% leads since January 2026
Not in model cards

Models answer expert financial questions grounded in business data. The score is average accuracy on eight offline tasks.

Online tasks have separate scoring. Official chart labels are hash-checked. Broad API names remain unlisted.

Released
January 2026
Kind
Finance
Level
University
Scored
Graded by a model
Answers
Short answer
Tasks
All public
Strongest model tested
GPT-5 #70 on the capability index
Last model added
December 2025
Scores refreshed
November 12th, 2026
21
Entries on its own leaderboard, 9 models we track

GPT-5 leads with 61.5% among models released by December 2025

Best score per entry from the benchmark's own leaderboard, read on November 12th, 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.

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:

21 entries. Scroll sideways for the rest.

  1. 61.5 GPT-5
  2. 61.3 Gemini-3 (variant unspecified)
  3. 59.4 Doubao-Seed-1.6 (snapshot unspecified)
  4. 53.3 Qwen3-235B-A22B-Thinking
  5. 41.8 Kimi-K2 (snapshot unspecified)
  6. 41.7 DeepSeek-V3.2
  7. 41.3 Qwen3-32B
  8. 39.5 Claude Sonnet 4
  9. 37.2 Grok 4
  10. 35.9 Qwen2.5-72B
  11. 35.7 Dianjin-R1
  12. 33.8 GLM-Z1-32B
  13. 31.5 GLM-Z1-9B
  14. 31.3 DeepSeek-R1-Distill-Qwen-32B
  15. 27.9 FinX1
  16. 21.8 Fino1
  17. 21.0 Qwen2.5-7B-Instruct
  18. 19.0 InternLM2.5-20B
  19. 14.3 InternLM2.5-7B
  20. 13.1 DeepSeek-R1-Distill-Qwen-7B
  21. 11.5 Fin-R1
  1. 1 GPT-5 61.5%
  2. 2 Gemini-3 (variant unspecified) 61.3%
  3. 3 Doubao-Seed-1.6 (snapshot unspecified) 59.4%
  4. 4 Qwen3-235B-A22B-Thinking Jul 2025 53.3%
  5. 5 Kimi-K2 (snapshot unspecified) 41.8%
  6. 6 DeepSeek-V3.2 41.7%
  7. 7 Qwen3-32B 41.3%
  8. 8 Claude Sonnet 4 39.5%
  9. 9 Grok 4 37.2%
  10. 10 Qwen2.5-72B 35.9%
  11. 11 Dianjin-R1 35.7%
  12. 12 GLM-Z1-32B 33.8%
  13. 13 GLM-Z1-9B 31.5%
  14. 14 DeepSeek-R1-Distill-Qwen-32B 31.3%
  15. 15 FinX1 27.9%
  16. 16 Fino1 21.8%
  17. 17 Qwen2.5-7B-Instruct 21%
  18. 18 InternLM2.5-20B 19%
  19. 19 InternLM2.5-7B 14.3%
  20. 20 DeepSeek-R1-Distill-Qwen-7B 13.1%
  21. 21 Fin-R1 11.5%

Source: Official BizFinBench v2 leaderboard

#EntryScoreSettingDateSource
1 GPT-5 61.5% Aug 2025 Official leaderboard
2 Gemini-3 (variant unspecified) 61.3% Official leaderboard
3 Doubao-Seed-1.6 (snapshot unspecified) 59.4% Official leaderboard
4 Qwen3-235B-A22B-Thinking Jul 2025 53.3% Jul 2025 Official leaderboard
5 Kimi-K2 (snapshot unspecified) 41.8% Official leaderboard
6 DeepSeek-V3.2 41.7% Dec 2025 Official leaderboard
7 Qwen3-32B 41.3% Apr 2025 Official leaderboard
8 Claude Sonnet 4 39.5% May 2025 Official leaderboard
9 Grok 4 37.2% Jul 2025 Official leaderboard
10 Qwen2.5-72B 35.9% Sep 2024 Official leaderboard
11 Dianjin-R1 35.7% Official leaderboard
12 GLM-Z1-32B 33.8% Official leaderboard
13 GLM-Z1-9B 31.5% Official leaderboard
14 DeepSeek-R1-Distill-Qwen-32B 31.3% Jan 2025 Official leaderboard
15 FinX1 27.9% Official leaderboard
16 Fino1 21.8% Official leaderboard
17 Qwen2.5-7B-Instruct 21% Official leaderboard
18 InternLM2.5-20B 19% Official leaderboard
19 InternLM2.5-7B 14.3% Official leaderboard
20 DeepSeek-R1-Distill-Qwen-7B 13.1% Jan 2025 Official leaderboard
21 Fin-R1 11.5% Official leaderboard

Number one over time

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

OpenAI
Own leaderboard Not tested since Jan 2026
GPT-5 61.5%

Source: Official BizFinBench v2 leaderboard

Scores over time

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

On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-5 Qwen3-235B-A22B-Thinking Qwen3-32B Qwen2.5-72B 0% 20% 40% 60% 80% 100% 2025 2026 Near the ceiling (85%) GPT-5 Qwen3-235B-A22B-Thinking

Source: Official BizFinBench v2 leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

Versions

We switch the top row to each new version when it comes out. Open it for one row per version. An old version stays in play while people still test it, so whoever first takes it past 85% gets credit for closing it out.

Qwen OpenAI
BizFinBench v2 Not tested since Jan 2026
GPT-5 61.5%
BizFinBench Not tested since May 2025
Qwen2.5-VL-32B 68.6%
BizFinBench v2 Not tested since Jan 2026
GPT-5 61.5%

Source: Official leaderboards

Replaces BizFinBench

More in Law, finance and medicine

All 179 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%
  • KBL In use Models answer Korean legal knowledge and reasoning tasks and Korean bar exam questions. It has 7 knowledge tasks, 4 reasoning tasks and 2,510 bar exam items. Solar Pro 4 77.5%
  • 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. Solar Pro 4 93.2%