# Hy4 Benchmark Scores & Performance

> Hy4 by Tencent has 15 source-displayable benchmark rows but no public overall score. It remains unranked.

Canonical page: https://benchlm.ai/models/hy4

Last updated: November 4, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | Tencent |
| Source Type | Pending |
| Reasoning Type | Reasoning |
| Context Window | 1M |
| Official model card | [Tencent Hy4 launch post](https://hy.tencent.ai/research/hy4) |
| Overall Score | Not computed (unranked) |
| Overall Rank | Unranked |

## Family & Coverage

- Family: Hy4
- Variant: base
- Benchmarks covered: 15 of 739
- Sibling models: [Hy4 preview](/models/hy4-preview)
- Related earlier model: [Hy4 preview](/models/hy4-preview)
- Coverage note: Published evidence is shown below; no public overall score is assigned.

## Agentic Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [Terminal-Bench 4.0](/benchmarks/terminal-bench-4) | 44.40% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [Terminal-Bench-Science 0.1](/benchmarks/terminal-bench-science) | 19.8% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [HLE w/ tools](/benchmarks/hlewithtools) | 59.1% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [Agents' Last Exam](/benchmarks/agentslastexam) | 41.0% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [APEX-Agents](/benchmarks/apexagents) | 75.2% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [AutomationBench](/benchmarks/automationbench) | 56.2% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [OSWorld 2.1](/benchmarks/osworld-2-1) | 78.9% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |

## Coding Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [DeepSWE](/benchmarks/deepswe) | 74.9% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [sweMarathon](/benchmarks/swemarathon) | 43.8% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [ProgramBench](/benchmarks/programbench) | 30.0% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [FrontierCode 2.0](/benchmarks/frontiercode-2) | 31.0% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [PostTrainBench v1.2](/benchmarks/posttrainbench12) | 35.2% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |

## Multimodal & Grounded Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [MMMU-Pro](/benchmarks/mmmu-pro) | 78.6% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |
| [ZeroBench w/ Python](/benchmarks/zerobenchpython) | 55.4% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |

## Knowledge Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [HLE](/benchmarks/hle) | 59.1% | [Tencent Hy4 launch post benchmark appendix](https://hy.tencent.ai/research/hy4) |

## Other Tencent Models

- [Hy4 preview](/models/hy4-preview) - Score: 61.46
- [Hy3 Preview](/models/hy3-preview) - Score: 52.95
- [Hy3](/models/hy3) - Score: not computed
- [UI-Mate-27B](/models/ui-mate-27b) - Score: not computed
- [UI-Mate-9B](/models/ui-mate-9b) - Score: not computed
- [Hy-Memory-30B-A3B](/models/hy-memory-30b-a3b) - Score: not computed
- [AuK](/models/tencent-auk) - Score: not computed
- [Hy ASR 3.0 preview](/models/hy-asr-3-0-preview) - Score: not computed
- [Hy-MT2-1.8B](/models/hy-mt2-1-8b) - Score: not computed
- [Hy-MT2-30B-A3B](/models/hy-mt2-30b-a3b) - Score: not computed