ChouBun
Models answer questions about long Japanese documents and summarize company securities reports.
The dataset card reports three question answering scores and one summarization score.
- Made by
- Sakana AI
- Released
- October 2024
- Kind
- Reasoning over long documents
- Level
- University
- Scored
- Checked automatically
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Subject
- Economics and business
- Language
- Japanese
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Law and finance
- Strongest model tested
-
Llama 3-8B #206 on the capability index
- Last model added
- April 2024
- Scores refreshed
- November 20th, 2026
elyza/Llama-3-ELYZA-JP-8B tops the leaderboard with 29.5 Llama 2-7B leads with 17.4 among models released by April 2024
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. The best model we track is Llama 2-7B, at 17.4. This view shows only models we track, each at its best.
The newest model with a score here came out in April 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 elyza/Llama-3-ELYZA-JP-8B Maximum input length 8192 29.5
- 2 01-ai/yi-6b-200k Maximum input length 200000 28.3
- 3
Llama 2-7B Maximum input length 2048 17.4
- 4
Llama 3-8B Maximum input length 8192 17.3
- 5 rinna/llama-3-youko-8b Maximum input length 8192 17.1
- 6
Mistral 7B v0.1 Maximum input length 32768 10.9
Source: Official ChouBun leaderboard
Source: Official ChouBun leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | elyza/Llama-3-ELYZA-JP-8B | 29.5 | Maximum input length 8192 | Official leaderboard | |
| 2 | 01-ai/yi-6b-200k | 28.3 | Maximum input length 200000 | Official leaderboard | |
| 3 | | 17.4 | Maximum input length 2048 | Jul 2023 | Official leaderboard |
| 4 | | 17.3 | Maximum input length 8192 | Apr 2024 | Official leaderboard |
| 5 | rinna/llama-3-youko-8b | 17.1 | Maximum input length 8192 | Official leaderboard | |
| 6 | | 10.9 | Maximum input length 32768 | Sep 2023 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 17.4 | Maximum input length 2048 | Jul 2023 | Official leaderboard |
| 2 | | 17.3 | Maximum input length 8192 | Apr 2024 | Official leaderboard |
| 3 | | 10.9 | Maximum input length 32768 | Sep 2023 | Official leaderboard |
Rankings by part
We read 4 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as wiki_qa, editnet_qa and corp_sec_qa. Llama 2-7B leads 2 of them, and Llama 3-8B leads the other 2. Arrows show how far a model moved from its overall rank among the models we track.
wiki_qa
3 models we track have a score here.
- 1
Llama 2-7B 16.8
- 2
Llama 3-8B 14.6
- 3
Mistral 7B v0.1 8.68
editnet_qa
3 models we track have a score here.
- 1
Llama 3-8B ↑1 14.8
- 2
Llama 2-7B ↓1 9.92
- 3
Mistral 7B v0.1 8.34
corp_sec_qa
3 models we track have a score here.
- 1
Llama 2-7B 20.9
- 2
Llama 3-8B 16.9
- 3
Mistral 7B v0.1 16.3
corp_sec_sum
3 models we track have a score here.
- 1
Llama 3-8B ↑1 22.8
- 2
Llama 2-7B ↓1 22.0
- 3
Mistral 7B v0.1 10.5
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official ChouBun leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official ChouBun 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 Long context
- MRCR v2 In use Multi-round coreference: retrieve the n-th of several similar requests hidden in a long synthetic conversation, up to 1M tokens.
Muse Spark 1.3 98.5%
- LongBench v2 In use 503 multiple-choice questions over 8K-2M-word contexts (QA, dialogue, code repos, structured data). Accuracy.
Inkling 2 Small 70.6%
- GraphWalks In use A graph is given as an edge list in long context. The model answers BFS or parent queries, scored by F1. Claude Fable 5 99.96%
- OpenAI MRCR v2 In use Models read a long synthetic chat that hides 2, 4 or 8 identical writing requests and must reproduce the response to the i-th one. Scored by mean string match ratio per context bin.
Gemini 3.7 Flash 89%
- CL-bench In use Model must learn new knowledge from a supplied long context (rules, procedures, data) and solve tasks. Strict LLM-verified rubrics. GPT-5.4 27.9%
- AA-LCR 1.1 In use Models answer 100 questions that need reasoning across sets of real documents of about 100K tokens. An LLM equality checker with a fixed system prompt compares each answer with the key.
Step 5 Preview 88.3%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in April 2024, for Llama 3-8B. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (6 entries).
- Pass Models scored 3 models we track, plus 3 other entries on its leaderboard.
- Warning Newest score The newest came 2 years ago, in April 2024, for Llama 3-8B.
- Note Frontier models None of the top 10 has a score. The strongest tested is Llama 3-8B (#206).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on huggingface.co: 3 rows for models we track and 3 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.