NusaWrites
Models interpret and translate texts in underrepresented Indonesian languages. Google reports emotion classification on original paragraphs and retrieval of matching translations.
- Made by
- NusaWrites team
- Released
- September 2023
- Kind
- Reading in many languages
- Level
- School
- Scored
- Checked automatically
- Answers
- Prediction
- Tasks
- All public
- Measures
- Capability
- Language
- Many languages
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Many languages
Logistic Regression tops the leaderboard with 78.2%
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The paper reports no overall score, so the headline is macro-F1 for NusaParagraph emotion classification. 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.
- 78.2 Logistic Regression
- 76.4 SVM
- 75.5 Naive Bayes
- 67.4 XLM-RLARGE
- 67.1 IndoNLU IndoBERTBASE
- 66.9 IndoLEM IndoBERTBASE
- 63.1 mBERT
- 62.6 IndoNLU IndoBERTLARGE
- 59.1 XLM-RBASE
- 22.0 mT0XL
- 19.1 mT0XXL
- 13.9 BLOOMZ-7.1B
- 13.8 mT0BASE
- 12.2 mT0LARGE
- 11.7 BLOOMZ-1.1B
- 9.35 mT0SMALL
- 8.56 BLOOMZ-1.7B
- 8.54 BLOOMZ-3B
- 6.57 BLOOMZ-560M
- 1 Logistic Regression Classical 78.2%
- 2 SVM Classical 76.4%
- 3 Naive Bayes Classical 75.5%
- 4 XLM-RLARGE Fine-tuning 67.4%
- 5 IndoNLU IndoBERTBASE Fine-tuning 67.1%
- 6 IndoLEM IndoBERTBASE Fine-tuning 66.9%
- 7 mBERT Fine-tuning 63.1%
- 8 IndoNLU IndoBERTLARGE Fine-tuning 62.6%
- 9 XLM-RBASE Fine-tuning 59.1%
- 10 mT0XL Zero-shot 22%
- 11 mT0XXL Zero-shot 19.1%
- 12 BLOOMZ-7.1B Zero-shot 13.9%
- 13 mT0BASE Zero-shot 13.8%
- 14 mT0LARGE Zero-shot 12.2%
- 15 BLOOMZ-1.1B Zero-shot 11.7%
- 16 mT0SMALL Zero-shot 9.35%
- 17 BLOOMZ-1.7B Zero-shot 8.56%
- 18 BLOOMZ-3B Zero-shot 8.54%
- 19 BLOOMZ-560M Zero-shot 6.57%
Source: The NusaWrites paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Logistic Regression | 78.2% | Classical | Official leaderboard | |
| 2 | SVM | 76.4% | Classical | Official leaderboard | |
| 3 | Naive Bayes | 75.5% | Classical | Official leaderboard | |
| 4 | XLM-RLARGE | 67.4% | Fine-tuning | Official leaderboard | |
| 5 | IndoNLU IndoBERTBASE | 67.1% | Fine-tuning | Official leaderboard | |
| 6 | IndoLEM IndoBERTBASE | 66.9% | Fine-tuning | Official leaderboard | |
| 7 | mBERT | 63.1% | Fine-tuning | Official leaderboard | |
| 8 | IndoNLU IndoBERTLARGE | 62.6% | Fine-tuning | Official leaderboard | |
| 9 | XLM-RBASE | 59.1% | Fine-tuning | Official leaderboard | |
| 10 | mT0XL | 22% | Zero-shot | Official leaderboard | |
| 11 | mT0XXL | 19.1% | Zero-shot | Official leaderboard | |
| 12 | BLOOMZ-7.1B | 13.9% | Zero-shot | Official leaderboard | |
| 13 | mT0BASE | 13.8% | Zero-shot | Official leaderboard | |
| 14 | mT0LARGE | 12.2% | Zero-shot | Official leaderboard | |
| 15 | BLOOMZ-1.1B | 11.7% | Zero-shot | Official leaderboard | |
| 16 | mT0SMALL | 9.35% | Zero-shot | Official leaderboard | |
| 17 | BLOOMZ-1.7B | 8.56% | Zero-shot | Official leaderboard | |
| 18 | BLOOMZ-3B | 8.54% | Zero-shot | Official leaderboard | |
| 19 | BLOOMZ-560M | 6.57% | Zero-shot | Official 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 Multilingual
- MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
- WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET.
K-EXAONE 90.5%
- MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language.
Qwen3.7-Max 87%
- CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories.
Solar Open 2 90.7%
- Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
- MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%
Data audit
All 8 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the benchmark’s paper (19 entries).
- Pass Models scored 19 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom No scores yet.
Leaderboard import 3 pass
- Pass Importer We read the board on aclanthology.org: 19 entries at the last read, none a model we track.
- 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.