HorizonMath (pass@4)
HorizonMath - long-horizon math reasoning, pass@4 protocol, reported in the Tencent Hy4 preview model card appendix.
Category
STEM & Reasoning
Score type
percent
Scale
100%
higher is better
Models tested
4
What is HorizonMath (pass@4)?
HorizonMath (pass@4) is a benchmark for hard reasoning in science, technology, engineering and math. The official description says: "HorizonMath - long-horizon math reasoning, pass@4 protocol, reported in the Tencent Hy4 preview model card appendix". Below we describe in simple words what benchmarks of this type generally examine.
What does it test?
HorizonMath (pass@4) measures hard reasoning in science, technology, engineering and math. In detail, it checks:
- graduate- and expert-level questions in science and math
- multi-step derivations where each step must be correct
- problems written to defeat memorisation (novel or recent exam problems)
- correct use of formulas, units and logical inference
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: HorizonMath - long-horizon math reasoning, pass@4 protocol, reported in the Tencent Hy4 preview model card appendix.
Frequently asked questions
What is the HorizonMath (pass@4) benchmark?
HorizonMath (pass@4) is a benchmark for hard reasoning in science, technology, engineering and math. The official description says: 'HorizonMath - long-horizon math reasoning, pass@4 protocol, reported in the Tencent Hy4 preview model card appendix'. Below we describe in simple words what benchmarks of this type generally examine.
What does HorizonMath (pass@4) test?
HorizonMath (pass@4) mainly measures hard reasoning in science, technology, engineering and math. Concretely, it checks: graduate- and expert-level questions in science and math; multi-step derivations where each step must be correct; problems written to defeat memorisation (novel or recent exam problems).
How is the HorizonMath (pass@4) 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 HorizonMath (pass@4) 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/horizonmath-pass4/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.
Benchmark dataset
The question/task data itself — download and evaluate locally
Benchmark metadata (JSON)
This page's benchmark record from the Open LLM Radar API
All model scores (JSON download)
Full score table as JSON; paginate with limit/offset params
Open LLM Radar API docs
Swagger UI to explore and reuse all radar API endpoints
Model scores 4 models tested
| # | Model | Score | Date | Source | Verified |
|---|---|---|---|---|---|
| 1 |
Hy4 preview
Tencent
|
8.8%
|
28.08.2026 | Tencent Hy4 preview model card benchmark appendix (https://huggingface.co/tencent/Hy4-preview#benchmark-appendix) | Verified |
| 2 |
Kimi K3
Moonshot AI
|
7.08%
|
28.08.2026 | Tencent Hy4 preview model card benchmark appendix (https://huggingface.co/tencent/Hy4-preview#benchmark-appendix) (Tencent rerun) | Verified |
| 3 |
Qwen3.8-2.4T-A95B
Qwen
|
5.31%
|
28.08.2026 | Tencent Hy4 preview model card benchmark appendix (https://huggingface.co/tencent/Hy4-preview#benchmark-appendix) (Tencent rerun) | Verified |
| 4 |
Hy3
Tencent
|
3.5%
|
28.08.2026 | Tencent Hy4 preview model card benchmark appendix (https://huggingface.co/tencent/Hy4-preview#benchmark-appendix) | Verified |