NOVA-63
NOVA-63 - multilingual benchmark covering 63 languages.
Category
Multilingual
Score type
percent
Scale
relative
higher is better
Models tested
5
What is NOVA-63?
NOVA-63 is a benchmark for performance in languages other than English. The official description says: "NOVA-63 - multilingual benchmark covering 63 languages". Below we describe in simple words what benchmarks of this type generally examine.
What does it test?
NOVA-63 measures performance in languages other than english. In detail, it checks:
- understanding and answering in many different languages
- translation quality between language pairs
- keeping math and reasoning quality when the question is not in English
- treating low-resource languages fairly instead of defaulting to English
How to read the score
The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better.
Official methodology note: NOVA-63 - multilingual benchmark covering 63 languages.
Frequently asked questions
What is the NOVA-63 benchmark?
NOVA-63 is a benchmark for performance in languages other than English. The official description says: 'NOVA-63 - multilingual benchmark covering 63 languages'. Below we describe in simple words what benchmarks of this type generally examine.
What does NOVA-63 test?
NOVA-63 mainly measures performance in languages other than English. Concretely, it checks: understanding and answering in many different languages; translation quality between language pairs; keeping math and reasoning quality when the question is not in English.
How is the NOVA-63 score calculated?
The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better.
Where do I get NOVA-63 data and scores?
The official benchmark page/reference is linked on this page. Every score on Open LLM Radar is also available as open data: download the scores JSON (http://open-llm-radar-api:8000/api/v1/benchmarks/nova-63/scores) or pull it directly from the Open LLM Radar API with any HTTP client.
Benchmark data & downloads
Where to get, inspect and reuse the data behind this benchmark.
Model scores 5 models tested
| # | Model | Score | Date | Source | Verified |
|---|---|---|---|---|---|
| 1 |
Qwen3.5-122B-A10B
Qwen Team (Alibaba Cloud)
|
58.6%
|
01.02.2026 | Report | Verified |
| 2 |
Qwen3.5-27B
Qwen Team (Alibaba Cloud)
|
58.1%
|
01.02.2026 | Report | Unverified |
| 3 |
Qwen3.5-35B-A3B
Qwen Team (Alibaba Cloud)
|
57.1%
|
01.02.2026 | Report | Unverified |
| 4 |
Qwen3-235B-A22B
Qwen Team (Alibaba Cloud)
|
55.4%
|
01.02.2026 | Report | Unverified |
| 5 |
Qwen3.5-4B
Qwen
|
54.3%
|
01.02.2026 | Report | Verified |