# Claude Opus 4.5 vs Hy3 Preview

> Side-by-side benchmark comparison for Claude Opus 4.5 and Hy3 Preview across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/claude-opus-4-5-vs-hy3-preview
- Last updated: September 23, 2026

- Shared sourced benchmarks: 4
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.6

## Quick Verdict

Pick Claude Opus 4.5 if you want the stronger benchmark profile. Hy3 Preview only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- Claude Opus 4.5 leads overall 55.9 to 45.32.
- The clearest category separation is in instruction following, where the averages are 30.6 for Claude Opus 4.5 and 46.9 for Hy3 Preview.
- The biggest single benchmark swing is Gert Labs in Agentic, with scores of 64.23% and 36.91%.
- Hy3 Preview is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- Hy3 Preview also has the larger context window at 256K.

## Model Snapshot

| Property | Claude Opus 4.5 | Hy3 Preview |
|----------|----------|----------|
| Creator | Anthropic | Tencent |
| Type | Proprietary | Open Weight |
| Reasoning | Non-Reasoning | Reasoning |
| Context | 200K | 256K |
| Overall Score | 55.9 | 45.32 |
| Benchmarks Covered | 45 | 6 |
| Pricing (input/output) | $5.00 / $25.00 | $0.00 / $0.00 |

## Category Breakdown

### Agentic

- Winner: Not comparable
- Claude Opus 4.5 public-lane score: 31.4 (Estimated · #64/105)
- Hy3 Preview public-lane score: Coming soon

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.0 | 59.3% | 54.4% | Claude Opus 4.5 |
| OSWorld-Verified | 66.3% | Coming soon | Coming soon |
| OSWorld | 66.3% | Coming soon | Coming soon |
| Claw-Eval | 59.6% | Coming soon | Coming soon |
| QwenClawBench | 52.3% | Coming soon | Coming soon |
| τ³-bench results | 70.2% | Coming soon | Coming soon |
| VITA-Bench | 23.3% | Coming soon | Coming soon |
| DeepPlanning | 26.4% | Coming soon | Coming soon |
| Toolathlon | 43.5% | Coming soon | Coming soon |
| MCP Atlas | 42.3% | Coming soon | Coming soon |
| MCP-Tasks | 71.8% | Coming soon | Coming soon |
| WideResearch | 76.4% | Coming soon | Coming soon |
| CyberGym | 50.6% | Coming soon | Coming soon |
| Gert Labs | 64.23% | 36.91% | Claude Opus 4.5 |
| JobBench | 32.3% | Coming soon | Coming soon |

### Coding

- Winner: Directional only
- Claude Opus 4.5 public-lane score: 46.1 (Estimated · #48/135)
- Hy3 Preview public-lane score: 33.3 (Estimated · #85/135)

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Verified | 80.9% | 74.4% | Claude Opus 4.5 |
| LiveCodeBench v6 | 84.8% | Coming soon | Coming soon |
| SWE-bench Pro | 57.1% | Coming soon | Coming soon |
| SWE Multilingual | 77.5% | Coming soon | Coming soon |
| NL2Repo | 43.2% | Coming soon | Coming soon |
| Terminal-Bench 2.0 | Coming soon | 54.4% | Coming soon |

### Reasoning

- Winner: Not comparable
- Claude Opus 4.5 public-lane score: 72.1 (Unranked · 4 rankable rows)
- Hy3 Preview public-lane score: Coming soon

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| LongBench v2 | 64.4% | Coming soon | Coming soon |
| AI-Needle | 74% | Coming soon | Coming soon |

### Multimodal & Grounded

- Winner: Not comparable
- Claude Opus 4.5 public-lane score: 23.5 (#48/50)
- Hy3 Preview public-lane score: Coming soon

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| MMMU-Pro | 70.6% | Coming soon | Coming soon |
| MathVision | 74.3% | Coming soon | Coming soon |
| CharXiv | 68.5% | Coming soon | Coming soon |
| VideoMMMU | 84.4% | Coming soon | Coming soon |
| ScreenSpot Pro | 45.7% | Coming soon | Coming soon |
| V* | 67.0% | Coming soon | Coming soon |

### Knowledge

- Winner: Directional only
- Claude Opus 4.5 public-lane score: 49 (Estimated · #57/160)
- Hy3 Preview public-lane score: 39.1 (Estimated · #93/160)

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 87% | 87.2% | Hy3 Preview |
| SuperGPQA | 70.6% | Coming soon | Coming soon |
| MMLU-Pro | 89.5% | Coming soon | Coming soon |
| MMLU-Redux | 96.6% | Coming soon | Coming soon |
| C-Eval | 92.2% | Coming soon | Coming soon |
| HLE | 30.8% | Coming soon | Coming soon |
| GPQA-D | Coming soon | 87.2% | Coming soon |

### Multilingual

- Winner: Not comparable
- Claude Opus 4.5 public-lane score: 82.9 (#2/12)
- Hy3 Preview public-lane score: Coming soon

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| MMLU-ProX | 85.7% | Coming soon | Coming soon |
| NOVA-63 | 56.7% | Coming soon | Coming soon |

### Instruction Following

- Winner: Directional only
- Claude Opus 4.5 public-lane score: 30.6 (#114/124)
- Hy3 Preview public-lane score: 46.9 (#86/124)

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| IFEval | 90.9% | Coming soon | Coming soon |
| IFBench | 58% | Coming soon | Coming soon |

### Mathematics

- Winner: Not comparable
- Claude Opus 4.5 public-lane score: 57.9 (#6/7)
- Hy3 Preview public-lane score: Coming soon

| Benchmark | Claude Opus 4.5 | Hy3 Preview | Winner |
|-----------|-----------|-----------|--------|
| AIME26 | 95.1% | Coming soon | Coming soon |
| HMMT Feb 2025 | 92.9% | Coming soon | Coming soon |
| HMMT Nov 2025 | 93.3% | Coming soon | Coming soon |
| HMMT Feb 2026 | 85.3% | Coming soon | Coming soon |
| MMAnswerBench | 84.0% | Coming soon | Coming soon |
| FrontierMath v2 (Tiers 1-3) | 20.690% | Coming soon | Coming soon |
| FrontierMath v2 (Tier 4) | 4.167% | Coming soon | Coming soon |

## FAQ

### Which is better overall, Claude Opus 4.5 or Hy3 Preview?

Claude Opus 4.5 is ahead overall on BenchLM right now.

### Where is the biggest gap between Claude Opus 4.5 and Hy3 Preview?

The widest category gap is in instruction following, where the averages are 30.6 for Claude Opus 4.5 and 46.9 for Hy3 Preview.

### How many benchmarks does Claude Opus 4.5 cover on BenchLM?

Claude Opus 4.5 currently has 45 sourced benchmark scores on BenchLM.

### How many benchmarks does Hy3 Preview cover on BenchLM?

Hy3 Preview currently has 6 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Claude Opus 4.5 vs Claude Opus 4.5 Thinking](/compare/claude-opus-4-5-vs-claude-opus-4-5-thinking)
- [Hy3 Preview vs Claude Opus 4.5 Thinking](/compare/claude-opus-4-5-thinking-vs-hy3-preview)
- [Claude Opus 4.5 vs Hy3](/compare/claude-opus-4-5-vs-hy3)
- [Hy3 Preview vs Hy3](/compare/hy3-vs-hy3-preview)

## Explore More

- [Claude Opus 4.5 profile](/models/claude-opus-4-5)
- [Hy3 Preview profile](/models/hy3-preview)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)