NewsBench (Chinese journalism)
Models complete Chinese newsroom assignments and answer multiple choice questions about writing and editorial safety. GPT-4 evaluates the written responses.
The paper compares 11 models on 1,267 items. The authors’ leaderboard adds Xinyu2-70B and reports four scores for each of 12 models.
- Made by
- Miao Li and colleagues at the Institute for Advanced Algorithms Research, Shanghai, and partner institutions
- Released
- February 2024
- Kind
- Other professions
- Level
- University
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- All public
- Measures
- Capability
- Subject
- Media and communications
- Language
- Chinese
- Interaction
- Single turn
- Items from
- Written by experts
- Refresh
- One fixed set
- Run by
- Anyone
- Occupations
- Arts, design, entertainment, sports and media
- Strongest model tested
- GPT-4 Turbo Nov 2023 #176 on the capability index
- Last model added
- November 2023
- Scores refreshed
- November 20th, 2026
GPT-4 Turbo leads with 2.44 among models released by November 2023 GPT-4 Turbo leads with 2.44 among models released by November 2023
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.
The newest model with a score here came out in November 2023, 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 GPT-4 Turbo Nov 2023 2.44
- 2 ERNIE Bot 2.41
- 3 Baichuan2-53B 2.41
- 4 Xverse 2.40
- 5
Qwen-14B 2.38
- 6 GPT-3.5-turbo (snapshot unspecified) 2.38
- 7
Baichuan2-13B 2.34
- 8 ChatGLM3-6B 2.31
- 9 Xinyu2-70B 2.29
- 10 chatglm2-6b 2.27
- 11 internlm-20b 2.22
- 12 Aquila-34B 2.18
Source: Official NewsBench (Chinese journalism) leaderboard
- 1 GPT-4 Turbo Nov 2023 2.44
- 2
Qwen-14B 2.38
- 3
Baichuan2-13B 2.34
- 4 ChatGLM3-6B 2.31
- 5 chatglm2-6b 2.27
- 6 internlm-20b 2.22
Source: Official NewsBench (Chinese journalism) leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-4 Turbo Nov 2023 | 2.44 | Nov 2023 | Official leaderboard | |
| 2 | ERNIE Bot closed weights | 2.41 | Official leaderboard | ||
| 3 | Baichuan2-53B open weights, 53B | 2.41 | Official leaderboard | ||
| 4 | Xverse open weights, 13B | 2.40 | Official leaderboard | ||
| 5 | | 2.38 | Sep 2023 | Official leaderboard | |
| 6 | GPT-3.5-turbo (snapshot unspecified) closed weights | 2.38 | Official leaderboard | ||
| 7 | | 2.34 | Sep 2023 | Official leaderboard | |
| 8 | ChatGLM3-6B | 2.31 | Oct 2023 | Official leaderboard | |
| 9 | Xinyu2-70B closed weights, 70B | 2.29 | Official leaderboard | ||
| 10 | chatglm2-6b | 2.27 | Jun 2023 | Official leaderboard | |
| 11 | internlm-20b | 2.22 | Sep 2023 | Official leaderboard | |
| 12 | Aquila-34B open weights, 34B | 2.18 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-4 Turbo Nov 2023 | 2.44 | Nov 2023 | Official leaderboard | |
| 2 | | 2.38 | Sep 2023 | Official leaderboard | |
| 3 | | 2.34 | Sep 2023 | Official leaderboard | |
| 4 | ChatGLM3-6B | 2.31 | Oct 2023 | Official leaderboard | |
| 5 | chatglm2-6b | 2.27 | Jun 2023 | Official leaderboard | |
| 6 | internlm-20b | 2.22 | 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 JWP-Generation, JWP-Multiple and SA-Generation. GPT-4 Turbo leads every one of them. Arrows show how far a model moved from its overall rank among the models we track.
JWP-Generation
6 models we track have a score here.
- 1 GPT-4 Turbo Nov 2023 2.44
- 2
Qwen-14B 2.38
- 3
Baichuan2-13B 2.34
- 4 ChatGLM3-6B 2.31
- 5 chatglm2-6b 2.27
- 6 internlm-20b 2.22
JWP-Multiple
6 models we track have a score here.
- 1 GPT-4 Turbo Nov 2023 0.46
- 2
Qwen-14B 0.44
- 3 internlm-20b ↑3 0.40
- 4
Baichuan2-13B ↓1 0.35
- 5 ChatGLM3-6B ↓1 0.33
- 6 chatglm2-6b ↓1 0.31
SA-Generation
6 models we track have a score here.
- 1 GPT-4 Turbo Nov 2023 0.90
- 2 internlm-20b ↑4 0.77
- 3 ChatGLM3-6B ↑1 0.76
- 4 chatglm2-6b ↑1 0.75
- 5
Baichuan2-13B ↓2 0.72
- 6
Qwen-14B ↓4 0.71
SA-Multiple
6 models we track have a score here.
- 1 GPT-4 Turbo Nov 2023 0.91
- 2
Qwen-14B 0.73
- 3
Baichuan2-13B 0.58
- 4 internlm-20b ↑2 0.58
- 5 chatglm2-6b 0.52
- 6 ChatGLM3-6B ↓2 0.49
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official NewsBench (Chinese journalism) 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 NewsBench (Chinese journalism) 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
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 3 years ago, in November 2023, for GPT-4 Turbo. 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 (12 entries).
- Pass Models scored 6 models we track, plus 6 other entries on its leaderboard.
- Warning Newest score The newest came 3 years ago, in November 2023, for GPT-4 Turbo.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-4 Turbo (#176).
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on iaar-shanghai.github.io: 6 rows for models we track and 6 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.