# Claude Haiku 4.5 vs GPT-5.5

> Side-by-side benchmark comparison for Claude Haiku 4.5 and GPT-5.5 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/claude-haiku-4-5-vs-gpt-5-5
- Last updated: September 4, 2026

- Shared sourced benchmarks: 8
- HTML indexing: indexable
- Ranking lane: BenchAlign v5

## Quick Verdict

Pick GPT-5.5 if you want the stronger benchmark profile. Claude Haiku 4.5 only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- GPT-5.5 leads overall 73.27 to 52.73.
- The clearest category separation is in mathematics, where the averages are 28.9 for Claude Haiku 4.5 and 69.6 for GPT-5.5.
- The biggest single benchmark swing is FrontierMath v2 (Tiers 1-3) in Mathematics, with scores of 5.903% and 51.700%.
- Claude Haiku 4.5 is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- GPT-5.5 also has the larger context window at 1M.

## Model Snapshot

| Property | Claude Haiku 4.5 | GPT-5.5 |
|----------|----------|----------|
| Creator | Anthropic | OpenAI |
| Type | Proprietary | Proprietary |
| Reasoning | Non-Reasoning | Reasoning |
| Context | 200K | 1M |
| Overall Score | 52.73 | 73.27 |
| Benchmarks Covered | 10 | 38 |
| Pricing (input/output) | $1.00 / $5.00 | $5.00 / $30.00 |

## Category Breakdown

### Agentic

- Winner: GPT-5.5
- Claude Haiku 4.5 public-lane score: 27 (Supported · #143/151)
- GPT-5.5 public-lane score: 63.9 (Supported · #15/151)

| Benchmark | Claude Haiku 4.5 | GPT-5.5 | Winner |
|-----------|-----------|-----------|--------|
| JobBench | 16.0% | 42.7% | GPT-5.5 |
| Terminal-Bench 2.1 (Vals) | 43.8% | 76.4% | GPT-5.5 |
| Terminal-Bench 2.0 | Coming soon | 82% | Coming soon |
| CyberGym | Coming soon | 81.8% | Coming soon |
| BrowseComp | Coming soon | 84.4% | Coming soon |
| OSWorld-Verified | Coming soon | 78.7% | Coming soon |
| MCP Atlas | Coming soon | 75.3% | Coming soon |
| Toolathlon | Coming soon | 55.6% | Coming soon |
| τ²-bench results | Coming soon | 98% | Coming soon |
| Gert Labs | Coming soon | 72.93% | Coming soon |
| ResearchClawBench | Coming soon | 17.0% | Coming soon |
| OSWorld 2.0 | Coming soon | 13.0% | Coming soon |
| ExploitGym | Coming soon | 13.4% | Coming soon |

### Coding

- Winner: GPT-5.5
- Claude Haiku 4.5 public-lane score: 27.3 (Supported · #173/183)
- GPT-5.5 public-lane score: 67.7 (Supported · #8/183)

| Benchmark | Claude Haiku 4.5 | GPT-5.5 | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Verified | 73.3% | Coming soon | Coming soon |
| VulcanBench v3 | 76.2% | Coming soon | Coming soon |
| LiveCodeBench (Vals) | 41.2% | 85.3% | GPT-5.5 |
| SWE-bench (Vals) | 66.6% | 82.6% | GPT-5.5 |
| SWE-bench Pro | Coming soon | 58.6% | Coming soon |
| Terminal-Bench 2.0 | Coming soon | 82.0% | Coming soon |
| Vibe Code Bench | Coming soon | 69.85% | Coming soon |
| React Native Evals | Coming soon | 84.7% | Coming soon |
| cursorBench31 | Coming soon | 59.2% | Coming soon |
| cursorBench32 | Coming soon | 58.4% | Coming soon |
| FrontierCode 1.1 Main | Coming soon | 43.0% | Coming soon |

### Multimodal & Grounded

- Winner: Coming soon
- Claude Haiku 4.5 public-lane score: Coming soon
- GPT-5.5 public-lane score: 71.3 (#19/48)

| Benchmark | Claude Haiku 4.5 | GPT-5.5 | Winner |
|-----------|-----------|-----------|--------|
| MMMU-Pro | Coming soon | 81.2% | Coming soon |
| MMMU-Pro w/ Python | Coming soon | 83.2% | Coming soon |
| OfficeQA Pro | Coming soon | 54.1% | Coming soon |

### Reasoning

- Winner: Coming soon
- Claude Haiku 4.5 public-lane score: Coming soon
- GPT-5.5 public-lane score: 63.5 (#15/22)

| Benchmark | Claude Haiku 4.5 | GPT-5.5 | Winner |
|-----------|-----------|-----------|--------|
| MRCR v2 64K-128K | Coming soon | 83.1% | Coming soon |
| MRCR v2 128K-256K | Coming soon | 87.5% | Coming soon |
| ARC-AGI-2 | Coming soon | 85% | Coming soon |
| ARC-AGI-3 | Coming soon | 0.4% | Coming soon |

### Knowledge

- Winner: Coming soon
- Claude Haiku 4.5 public-lane score: 44.6 (Estimated · #117/181)
- GPT-5.5 public-lane score: 73.3 (Supported · #7/181)

| Benchmark | Claude Haiku 4.5 | GPT-5.5 | Winner |
|-----------|-----------|-----------|--------|
| GPQA Diamond (Vals) | 72.2% | 93.2% | GPT-5.5 |
| MMLU-Pro (Vals) | 78.7% | 88.1% | GPT-5.5 |
| GPQA | Coming soon | 93.6% | Coming soon |
| GPQA-D | Coming soon | 93.6% | Coming soon |
| HLE | Coming soon | 52.2% | Coming soon |
| HLE w/o tools | Coming soon | 41.4% | Coming soon |

### Instruction Following

- Winner: Coming soon
- Claude Haiku 4.5 public-lane score: Coming soon
- GPT-5.5 public-lane score: 92.9 (#7/120)

### Mathematics

- Winner: Coming soon
- Claude Haiku 4.5 public-lane score: 28.9 (Unranked · 2 rankable rows)
- GPT-5.5 public-lane score: 69.6 (Unranked · 3 rankable rows)

| Benchmark | Claude Haiku 4.5 | GPT-5.5 | Winner |
|-----------|-----------|-----------|--------|
| FrontierMath v2 (Tiers 1-3) | 5.903% | 51.700% | GPT-5.5 |
| FrontierMath v2 (Tier 4) | 2.083% | 35.400% | GPT-5.5 |
| FrontierMath (legacy) | Coming soon | 51.7% | Coming soon |

## FAQ

### Which is better overall, Claude Haiku 4.5 or GPT-5.5?

GPT-5.5 is ahead overall on BenchLM right now.

### Where is the biggest gap between Claude Haiku 4.5 and GPT-5.5?

The widest category gap is in mathematics, where the averages are 28.9 for Claude Haiku 4.5 and 69.6 for GPT-5.5.

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

Claude Haiku 4.5 currently has 10 sourced benchmark scores on BenchLM.

### How many benchmarks does GPT-5.5 cover on BenchLM?

GPT-5.5 currently has 38 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Claude Haiku 4.5 vs Claude Haiku 4.5 Thinking](/compare/claude-haiku-4-5-vs-claude-haiku-4-5-thinking)
- [GPT-5.5 vs Claude Haiku 4.5 Thinking](/compare/claude-haiku-4-5-thinking-vs-gpt-5-5)
- [Claude Haiku 4.5 vs GPT-5.5 Pro](/compare/claude-haiku-4-5-vs-gpt-5-5-pro)
- [GPT-5.5 vs GPT-5.5 Pro](/compare/gpt-5-5-vs-gpt-5-5-pro)

## Explore More

- [Claude Haiku 4.5 profile](/models/claude-haiku-4-5)
- [GPT-5.5 profile](/models/gpt-5-5)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)