Knowledge Benchmark

OmniScience Accuracy

OmniScience Accuracy - scientific knowledge accuracy benchmark.

Category

Knowledge

Score type

percent

Scale

100%

higher is better

Models tested

7

What is OmniScience Accuracy?

OmniScience Accuracy is a benchmark for broad factual and academic knowledge. The official description says: "OmniScience Accuracy - scientific knowledge accuracy benchmark". Below we describe in simple words what benchmarks of this type generally examine.

What does it test?

OmniScience Accuracy measures broad factual and academic knowledge. In detail, it checks:

  • facts and concepts from many academic fields at once
  • harder variants with several plausible answer options
  • knowledge that goes beyond the model's training-set surface forms
  • resisting the plausible-but-wrong answers deliberately included

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: OmniScience Accuracy - scientific knowledge accuracy benchmark.

Frequently asked questions

What is the OmniScience Accuracy benchmark?

OmniScience Accuracy is a benchmark for broad factual and academic knowledge. The official description says: 'OmniScience Accuracy - scientific knowledge accuracy benchmark'. Below we describe in simple words what benchmarks of this type generally examine.

What does OmniScience Accuracy test?

OmniScience Accuracy mainly measures broad factual and academic knowledge. Concretely, it checks: facts and concepts from many academic fields at once; harder variants with several plausible answer options; knowledge that goes beyond the model's training-set surface forms.

How is the OmniScience Accuracy score calculated?

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better. The declared maximum score of this benchmark is 100.0.

Where do I get OmniScience Accuracy 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/omniscience-accuracy/scores) or pull it directly from the Open LLM Radar API with any HTTP client.

Model scores 7 models tested

Reset
# Model Score Date Source Verified
1 DeepSeek-V4-Pro
DeepSeek
46.8%
04.06.2026 Report Unverified
2 DeepSeek-V4-Flash
DeepSeek
39.9%
04.06.2026 Report Unverified
3 Kimi K2.6
Moonshot AI
35.5%
04.06.2026 Report Unverified
4 NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16
NVIDIA
24.1%
04.06.2026 Report Verified
5 NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB
NVIDIA
16.9%
15.10.2026 Report Verified
6 Kolibri-1
Aleph Alpha
14.8%
03.10.2026 Report Verified
7 LFM3-1.7B
LiquidAI
7.07%
— Report Verified

What now?

You've seen how models score on OmniScience Accuracy — here's what to do next.