# Step 5 Preview vs SWE-2

> Side-by-side benchmark comparison for Step 5 Preview and SWE-2 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/step-5-preview-vs-swe-2
- Last updated: September 21, 2026

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

## Quick Verdict

Pick Step 5 Preview if you want the stronger benchmark profile. SWE-2 only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- Step 5 Preview leads overall 72 to 0.
- The biggest single benchmark swing is Terminal-Bench 2.1 in Agentic, with scores of 85.0% and 92.8%.
- SWE-2 is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.

## Model Snapshot

| Property | Step 5 Preview | SWE-2 |
|----------|----------|----------|
| Creator | StepFun | Cognition |
| Type | Pending | Proprietary |
| Reasoning | Reasoning | Reasoning |
| Context | 1M | 1M |
| Overall Score | null | null |
| Benchmarks Covered | 22 | 5 |
| Pricing (input/output) | $1.00 / $2.70 | Pricing unavailable / Pricing unavailable |

## Category Breakdown

### Agentic

- Winner: Coming soon
- Step 5 Preview public-lane score: Coming soon
- SWE-2 public-lane score: Coming soon

| Benchmark | Step 5 Preview | SWE-2 | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.1 | 85.0% | 92.8% | SWE-2 |
| Terminal-Bench 4.0 | 33.30% | 27.30% | Step 5 Preview |
| CyberGym | 84.7% | Coming soon | Coming soon |
| AutomationBench | 44.0% | Coming soon | Coming soon |
| Toolathlon-Verified | 74.1% | Coming soon | Coming soon |
| MCP Atlas | 85.6% | Coming soon | Coming soon |
| JobBench | 59.0% | Coming soon | Coming soon |
| APEX-Agents | 37.8% | Coming soon | Coming soon |
| DRACO | 83.3% | Coming soon | Coming soon |
| BrowseComp | 88.7% | Coming soon | Coming soon |
| Agents' Last Exam | 29.5% | Coming soon | Coming soon |

### Coding

- Winner: Coming soon
- Step 5 Preview public-lane score: Coming soon
- SWE-2 public-lane score: Coming soon

| Benchmark | Step 5 Preview | SWE-2 | Winner |
|-----------|-----------|-----------|--------|
| DeepSWE | 67.7% | 73.0% | SWE-2 |
| Terminal-Bench 2.1 | 85.0% | 92.8% | SWE-2 |
| SciCode | 58.9% | Coming soon | Coming soon |
| ProgramBench | 80.5% | Coming soon | Coming soon |
| sweMarathon | 72.7% | Coming soon | Coming soon |
| MLS-Bench Lite | 40.5% | Coming soon | Coming soon |
| FrontierCode 1.1 Main | Coming soon | 50.0% | Coming soon |

### Reasoning

- Winner: Coming soon
- Step 5 Preview public-lane score: 78.5 (Unranked · 2 rankable rows)
- SWE-2 public-lane score: Coming soon

| Benchmark | Step 5 Preview | SWE-2 | Winner |
|-----------|-----------|-----------|--------|
| CritPt | 20.9% | Coming soon | Coming soon |

### Multimodal & Grounded

- Winner: Coming soon
- Step 5 Preview public-lane score: 61.2 (#28/49)
- SWE-2 public-lane score: Coming soon

| Benchmark | Step 5 Preview | SWE-2 | Winner |
|-----------|-----------|-----------|--------|
| OfficeQA Pro | 60.3% | Coming soon | Coming soon |
| MMMU-Pro | 76% | Coming soon | Coming soon |

### Knowledge

- Winner: Coming soon
- Step 5 Preview public-lane score: Coming soon
- SWE-2 public-lane score: Coming soon

| Benchmark | Step 5 Preview | SWE-2 | Winner |
|-----------|-----------|-----------|--------|
| GPQA-D | 93.5% | Coming soon | Coming soon |
| HLE | 46.5% | Coming soon | Coming soon |

## FAQ

### Which is better overall, Step 5 Preview or SWE-2?

Step 5 Preview is ahead overall on BenchLM right now.

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

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

### How many benchmarks does SWE-2 cover on BenchLM?

SWE-2 currently has 5 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Step 5 Preview vs SWE-1.7](/compare/step-5-preview-vs-swe-1-7)
- [SWE-2 vs SWE-1.7](/compare/swe-1-7-vs-swe-2)
- [Step 5 Preview vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-step-5-preview)
- [SWE-2 vs Claude Fable 5.1](/compare/claude-fable-5-1-vs-swe-2)

## Explore More

- [Step 5 Preview profile](/models/step-5-preview)
- [SWE-2 profile](/models/swe-2)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)