# Claude Sonnet 5.5 vs Step 5 Preview

> Side-by-side benchmark comparison for Claude Sonnet 5.5 and Step 5 Preview across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/claude-sonnet-5-5-vs-step-5-preview
- Last updated: September 28, 2026

- Shared sourced benchmarks: 6
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.7

## Quick Verdict

Pick Claude Sonnet 5.5 if you want the stronger benchmark profile. Step 5 Preview only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- Claude Sonnet 5.5 leads overall 80.49 to 72.
- The clearest category separation is in multimodal & grounded, where the averages are 83.3 for Claude Sonnet 5.5 and 61.2 for Step 5 Preview.
- The biggest single benchmark swing is Terminal-Bench 4.0 in Agentic, with scores of 70.60% and 33.30%.
- Step 5 Preview is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.

## Model Snapshot

| Property | Claude Sonnet 5.5 | Step 5 Preview |
|----------|----------|----------|
| Creator | Anthropic | StepFun |
| Type | Proprietary | Pending |
| Reasoning | Reasoning | Reasoning |
| Context | 1M | 1M |
| Overall Score | 80.49 | null |
| Benchmarks Covered | 47 | 22 |
| Pricing (input/output) | $2.00 / $10.00 | $1.00 / $2.70 |

## Category Breakdown

### Agentic

- Winner: Not comparable
- Claude Sonnet 5.5 public-lane score: 66.1 (Supported · #10/111)
- Step 5 Preview public-lane score: Coming soon

| Benchmark | Claude Sonnet 5.5 | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 4.0 | 70.60% | 33.30% | Claude Sonnet 5.5 |
| HLE w/ tools | 64.5% | Coming soon | Coming soon |
| DRACO | 87.0% | 83.3% | Claude Sonnet 5.5 |
| Terminal-Bench-Science 0.1 | 59.9% | Coming soon | Coming soon |
| LAB all-pass (Harvey held-out) | 10.0% | Coming soon | Coming soon |
| LAB criterion-pass (Harvey held-out) | 93.1% | Coming soon | Coming soon |
| Toolathlon Verified Pass@3 | 85.2% | Coming soon | Coming soon |
| Toolathlon Verified Pass³ | 68.5% | Coming soon | Coming soon |
| Toolathlon Verified avg. turns | 31.6 turns | Coming soon | Coming soon |
| AutomationBench (Zapier 1.0.6) | 44.7% | Coming soon | Coming soon |
| Toolathlon-Verified | 77.8% | 74.1% | Claude Sonnet 5.5 |
| Terminal-Bench 2.1 | Coming soon | 85.0% | Coming soon |
| CyberGym | Coming soon | 84.7% | Coming soon |
| AutomationBench | Coming soon | 44.0% | Coming soon |
| MCP Atlas | Coming soon | 85.6% | Coming soon |
| JobBench | Coming soon | 59.0% | Coming soon |
| APEX-Agents | Coming soon | 37.8% | Coming soon |
| BrowseComp | Coming soon | 88.7% | Coming soon |
| Agents' Last Exam | Coming soon | 29.5% | Coming soon |

### Coding

- Winner: Not comparable
- Claude Sonnet 5.5 public-lane score: 79.6 (Estimated · #3/136)
- Step 5 Preview public-lane score: Coming soon

| Benchmark | Claude Sonnet 5.5 | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| FrontierCode 1.1 Main | 46.2% | Coming soon | Coming soon |
| cursorBench40 | 55.5% | Coming soon | Coming soon |
| SWE-bench Pro | 81.3% | Coming soon | Coming soon |
| SWE Multilingual | 90.3% | Coming soon | Coming soon |
| SWE Multimodal | 54.3% | Coming soon | Coming soon |
| DeepSWE | 71.0% | 67.7% | Claude Sonnet 5.5 |
| FrontierCode 1.1 Extended | 59.1% | Coming soon | Coming soon |
| ProgramBench | 79.7% | 80.5% | Step 5 Preview |
| FrontierSWE v2 | 61.9% | Coming soon | Coming soon |
| Terminal-Bench 2.1 | Coming soon | 85.0% | Coming soon |
| SciCode | Coming soon | 58.9% | Coming soon |
| sweMarathon | Coming soon | 72.7% | Coming soon |
| MLS-Bench Lite | Coming soon | 40.5% | Coming soon |

### Reasoning

- Winner: Not comparable
- Claude Sonnet 5.5 public-lane score: 79.2 (#7/27)
- Step 5 Preview public-lane score: 79.6 (Unranked · 2 rankable rows)

| Benchmark | Claude Sonnet 5.5 | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| CritPt | Coming soon | 20.9% | Coming soon |

### Multimodal & Grounded

- Winner: Not comparable
- Claude Sonnet 5.5 public-lane score: 83.3 (Unranked · 7 rankable rows)
- Step 5 Preview public-lane score: 61.2 (#29/50)

| Benchmark | Claude Sonnet 5.5 | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| Chartography (no tools) | 61.6% | Coming soon | Coming soon |
| Chartography (tools) | 90.2% | Coming soon | Coming soon |
| BenchCAD Vision2Code (no tools) | 0.747 | Coming soon | Coming soon |
| BenchCAD Vision2Code (tools) | 0.963 | Coming soon | Coming soon |
| Biomedical image analysis | 72.2% | Coming soon | Coming soon |
| OfficeQA | 76.9% | Coming soon | Coming soon |
| OfficeQA Pro | 65.6% | 60.3% | Claude Sonnet 5.5 |
| MMMU-Pro | Coming soon | 76% | Coming soon |

### Knowledge

- Winner: Not comparable
- Claude Sonnet 5.5 public-lane score: 80.9 (Supported · #6/160)
- Step 5 Preview public-lane score: Coming soon

| Benchmark | Claude Sonnet 5.5 | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| HealthBench (raw) | 69.4% | Coming soon | Coming soon |
| HealthBench (length-adjusted) | 65.4% | Coming soon | Coming soon |
| HealthBench Professional (raw) | 77.1% | Coming soon | Coming soon |
| HealthBench Professional | 69.2% | Coming soon | Coming soon |
| BioMysteryBench (human-solvable) | 89.2% | Coming soon | Coming soon |
| BioMysteryBench (human-difficult) | 44.7% | Coming soon | Coming soon |
| SpatialBench Verified | 72.5% | Coming soon | Coming soon |
| SingleCellBench | 59.1% | Coming soon | Coming soon |
| Protein Design | 51.0% | Coming soon | Coming soon |
| Morphology-to-molecule matching | 25.0% | Coming soon | Coming soon |
| Medicinal chemistry | 65.3% | Coming soon | Coming soon |
| Protein Design library ranking | 54.8% | Coming soon | Coming soon |
| De novo protein-binder design | 82.3% | Coming soon | Coming soon |
| Protocols (troubleshooting) | 67.3% | Coming soon | Coming soon |
| Protocols (understanding) | 66.6% | Coming soon | Coming soon |
| HLE w/o tools | 56.9% | Coming soon | Coming soon |
| GPQA-D | Coming soon | 93.5% | Coming soon |
| HLE | Coming soon | 46.5% | Coming soon |

### Multilingual

- Winner: Not comparable
- Claude Sonnet 5.5 public-lane score: Coming soon
- Step 5 Preview public-lane score: Coming soon

| Benchmark | Claude Sonnet 5.5 | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| GMMLU | 92.1% | Coming soon | Coming soon |
| MILU | 91.6% | Coming soon | Coming soon |

### Mathematics

- Winner: Not comparable
- Claude Sonnet 5.5 public-lane score: Coming soon
- Step 5 Preview public-lane score: Coming soon

| Benchmark | Claude Sonnet 5.5 | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| ArXivMath Aug. 2026 (no tools) | 86.8% | Coming soon | Coming soon |
| ArXivMath Aug. 2026 (tools) | 95.2% | Coming soon | Coming soon |

## FAQ

### Which is better overall, Claude Sonnet 5.5 or Step 5 Preview?

Claude Sonnet 5.5 is ahead overall on BenchLM right now.

### Where is the biggest gap between Claude Sonnet 5.5 and Step 5 Preview?

The widest category gap is in multimodal & grounded, where the averages are 83.3 for Claude Sonnet 5.5 and 61.2 for Step 5 Preview.

### How many benchmarks does Claude Sonnet 5.5 cover on BenchLM?

Claude Sonnet 5.5 currently has 47 sourced benchmark scores on BenchLM.

### How many benchmarks does Step 5 Preview cover on BenchLM?

Step 5 Preview currently has 22 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Claude Sonnet 5.5 vs GPT-6 Astra](/compare/claude-sonnet-5-5-vs-gpt-6-astra)
- [Step 5 Preview vs GPT-6 Astra](/compare/gpt-6-astra-vs-step-5-preview)
- [Claude Sonnet 5.5 vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-claude-sonnet-5-5)
- [Step 5 Preview vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-step-5-preview)

## Explore More

- [Claude Sonnet 5.5 profile](/models/claude-sonnet-5-5)
- [Step 5 Preview profile](/models/step-5-preview)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)