JFinBench
Models answer questions from six Japanese certification exams in accounting, tax, financial planning, and banking. The authors report scores by exam.
- Made by
- Sakana AI
- Released
- November 2026
- Kind
- Finance
- Level
- University
- Scored
- Checked automatically
- Answers
- Multiple choice
- Tasks
- All public
- Measures
- Capability
- Subject
- Economics and business, Law
- Language
- Japanese
- Interaction
- Single turn
- Items from
- Exams and contests
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Law and finance
- Strongest model tested
- GPT-6 Astra #4 on the capability index
- Last model added
- October 2026
- Scores refreshed
- November 20th, 2026
GPT-6 Astra leads with 97.3%
Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.
- 1 GPT-6 Astra max, with tools 97.3%
- 2 Claude Fable 5 max, with tools 96.4%
- 3
Gemini 3.8 Flash max, with tools 94.2%
- 4 Kimi K3.1 Pro max, with tools 92.1%
- 5
DeepSeek V4.1 Flash 1017 max, with tools 88.6%
- 6 GLM-5.3-Flash max, with tools 87.6%
- 7 Kimi K3 max, with tools 86.1%
- 8 MiMo-V2.6 Pro thinking, with tools 85.3%
- 9
Qwen3.8 Flash default, with tools 83.3%
- 10 MiMo-V2.6 Flash thinking, with tools 73.2%
- 11 Kimi K2.6 thinking, with tools 70.4%
- 12
Mistral Large 4 thinking, with tools 70.1%
Source: Official JFinBench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-6 Astra | 97.3% | max, with tools | Sep 2026 | Official leaderboard |
| 2 | Claude Fable 5 | 96.4% | max, with tools | Jun 2026 | Official leaderboard |
| 3 | | 94.2% | max, with tools | Sep 2026 | Official leaderboard |
| 4 | Kimi K3.1 Pro | 92.1% | max, with tools | Oct 2026 | Official leaderboard |
| 5 | | 88.6% | max, with tools | Oct 2026 | Official leaderboard |
| 6 | GLM-5.3-Flash | 87.6% | max, with tools | Aug 2026 | Official leaderboard |
| 7 | Kimi K3 | 86.1% | max, with tools | Jul 2026 | Official leaderboard |
| 8 | MiMo-V2.6 Pro | 85.3% | thinking, with tools | Sep 2026 | Official leaderboard |
| 9 | | 83.3% | default, with tools | Aug 2026 | Official leaderboard |
| 10 | MiMo-V2.6 Flash | 73.2% | thinking, with tools | Sep 2026 | Official leaderboard |
| 11 | Kimi K2.6 | 70.4% | thinking, with tools | Apr 2026 | Official leaderboard |
| 12 | | 70.1% | thinking, with tools | Oct 2026 | Official leaderboard |
Rankings by part
We read 6 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as CPA, Tax accountant and FP. GPT-6 Astra leads 3 of them, and Claude Fable 5 leads the other 3. Arrows show how far a model moved from its overall rank among the models we track.
CPA
12 models we track have a score here.
- 1 GPT-6 Astra 99.6%
- 2 Claude Fable 5 98.6%
- 3
Gemini 3.8 Flash 97.8%
- 4 Kimi K3.1 Pro 96.1%
- 5
DeepSeek V4.1 Flash 1017 94.6%
- 6 GLM-5.3-Flash 93.5%
- 7 Kimi K3 91%
- 8 MiMo-V2.6 Pro 91%
- 9
Qwen3.8 Flash 89.6%
- 10 MiMo-V2.6 Flash 83.2%
Tax accountant
12 models we track have a score here.
- 1 GPT-6 Astra 87.2%
- 2 Claude Fable 5 87%
- 3
Gemini 3.8 Flash 78.2%
- 4 Kimi K3.1 Pro 72.7%
- 5 GLM-5.3-Flash ↑1 65.1%
- 6 Kimi K3 ↑1 64.1%
- 7
Qwen3.8 Flash ↑2 62.6%
- 8 MiMo-V2.6 Pro 61.2%
- 9
DeepSeek V4.1 Flash 1017 ↓4 58.6%
- 10 MiMo-V2.6 Flash 52.2%
FP
12 models we track have a score here.
- 1 Claude Fable 5 ↑1 100%
- 2 GPT-6 Astra ↓1 100%
- 3
Gemini 3.8 Flash 98.5%
- 4
DeepSeek V4.1 Flash 1017 ↑1 95.4%
- 5 Kimi K3.1 Pro ↓1 94.9%
- 6 Kimi K3 ↑1 91.8%
- 7 GLM-5.3-Flash ↓1 87.7%
- 8
Qwen3.8 Flash ↑1 85.1%
- 9 MiMo-V2.6 Pro ↓1 84.6%
- 10 MiMo-V2.6 Flash 73.8%
Counter services
12 models we track have a score here.
- 1 Claude Fable 5 ↑1 100%
- 2
Gemini 3.8 Flash ↑1 100%
- 3 GPT-6 Astra ↓2 100%
- 4 Kimi K3.1 Pro 99.5%
- 5 MiMo-V2.6 Pro ↑3 98%
- 6 Kimi K3 ↑1 97.5%
- 7
DeepSeek V4.1 Flash 1017 ↓2 96.5%
- 8 GLM-5.3-Flash ↓2 95.5%
- 9
Qwen3.8 Flash 91.4%
- 10
Mistral Large 4 ↑2 87.4%
Nissho Bookkeeping
12 models we track have a score here.
- 1 Claude Fable 5 ↑1 100%
- 2
Gemini 3.8 Flash ↑1 99.8%
- 3 GPT-6 Astra ↓2 98.4%
- 4 GLM-5.3-Flash ↑2 97%
- 5 Kimi K3.1 Pro ↓1 96.8%
- 6
DeepSeek V4.1 Flash 1017 ↓1 95.7%
- 7 Kimi K3 92.6%
- 8
Qwen3.8 Flash ↑1 90.1%
- 9 MiMo-V2.6 Pro ↓1 86.8%
- 10 Kimi K2.6 ↑1 84.6%
Zensho Bookkeeping
12 models we track have a score here.
- 1 GPT-6 Astra 99%
- 2 Claude Fable 5 98%
- 3 Kimi K3.1 Pro ↑1 94.9%
- 4 Kimi K3 ↑3 92.9%
- 5
Gemini 3.8 Flash ↓2 90.9%
- 6
DeepSeek V4.1 Flash 1017 ↓1 90.9%
- 7 MiMo-V2.6 Pro ↑1 89.9%
- 8 GLM-5.3-Flash ↓2 86.9%
- 9
Qwen3.8 Flash 80.8%
- 10
Mistral Large 4 ↑2 80.8%
The best score, 97.3%, costs $1.03 a task
Runs on the cost frontier score higher than any run that's cheaper. Claude Fable 5 at low comes within 4.7 points of the best for $0.06 a task, 17 times cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.
Source: Official JFinBench leaderboard
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| GPT-6 Astra | max, with tools | 97.3% | $1.03 | 13,780 | |
| Claude Fable 5 | max, with tools | 96.4% | $2.26 | 36,300 | |
| Claude Fable 5 | max | 95.8% | $0.56 | 10,854 | |
| max, with tools | 94.2% | $0.38 | 28,200 | ||
| GPT-6 Astra | max | 93.6% | $0.25 | 4,751 | |
| Claude Fable 5 | low | 92.6% | $0.06 | 938 | |
| GPT-6 Astra | low | 92.1% | $0.04 | 682 | |
| Kimi K3.1 Pro | max, with tools | 92.1% | $0.49 | 22,540 | |
| max | 88.8% | $0.03 | 6,567 | ||
| max, with tools | 88.6% | $0.05 | 133,020 | ||
| GLM-5.3-Flash | max, with tools | 87.6% | $0.22 | 132,080 | |
| Kimi K3 | max, with tools | 86.1% | $0.34 | 15,310 | |
| MiMo-V2.6 Pro | thinking, with tools | 85.3% | $0.09 | 82,180 | |
| low | 85.1% | $0.005 | 1,146 | ||
| Kimi K3.1 Pro | max | 83.4% | $0.15 | 14,262 | |
| default, with tools | 83.3% | $0.09 | 136,760 | ||
| Kimi K3 | max | 79.9% | $0.25 | 16,636 | |
| Kimi K3.1 Pro | low | 78.4% | $0.04 | 3,928 | |
| MiMo-V2.6 Flash | thinking, with tools | 73.2% | $0.04 | 106,570 | |
| max | 73.1% | $0.007 | 23,495 | ||
| Kimi K3 | low | 72.5% | $0.03 | 1,794 | |
| low | 71.7% | $0.003 | 8,953 | ||
| Kimi K2.6 | thinking, with tools | 70.4% | $0.35 | 50,580 | |
| thinking, with tools | 70.1% | $0.36 | 83,020 | ||
| default | 62.5% | $0.02 | 43,206 | ||
| MiMo-V2.6 Pro | thinking | 62.1% | $0.03 | 33,863 | |
| Kimi K2.6 | thinking | 59.1% | $0.06 | 14,162 | |
| thinking | 58.2% | $0.06 | 29,044 | ||
| GLM-5.3-Flash | max | 57.7% | $0.01 | 26,430 | |
| MiMo-V2.6 Flash | thinking | 50.9% | $0.01 | 34,421 | |
| GLM-5.3-Flash | low | 50.8% | $0.002 | 3,274 |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official JFinBench leaderboard
Claude Fable 5 on the official leaderboard passed 85% in June 2026
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.
Source: Official JFinBench 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
- 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 9 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (12 entries).
- Pass Models scored 12 models we track.
- Pass Newest score The newest is from October 2026, for Kimi K3.1 Pro.
- Note Frontier models 1 of the top 10 has a score, which is fine for a retired test. The strongest tested is GPT-6 Astra (#4).
- Note Headroom Saturated. The best is 97.3%, past 85% since June 2026.
Leaderboard import 3 pass
- Pass Importer We read the board on pub.sakana.ai: 31 rows for models we track 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.