# GPT-5.6 Terra vs Step 5 Preview

> Side-by-side benchmark comparison for GPT-5.6 Terra and Step 5 Preview across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/gpt-5-6-terra-vs-step-5-preview
- Last updated: September 21, 2026

- Shared sourced benchmarks: 5
- HTML indexing: indexable
- Ranking lane: BenchAlign v5

## Quick Verdict

Pick Step 5 Preview if you want the stronger benchmark profile. GPT-5.6 Terra 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 71.33.
- The clearest category separation is in multimodal & grounded, where the averages are 77.2 for GPT-5.6 Terra and 61.2 for Step 5 Preview.
- The biggest single benchmark swing is MMMU-Pro in Multimodal & Grounded, with scores of 80.7% and 76%.
- Step 5 Preview is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- GPT-5.6 Terra also has the larger context window at 1.05M.

## Model Snapshot

| Property | GPT-5.6 Terra | Step 5 Preview |
|----------|----------|----------|
| Creator | OpenAI | StepFun |
| Type | Proprietary | Pending |
| Reasoning | Reasoning | Reasoning |
| Context | 1.05M | 1M |
| Overall Score | 71.33 | null |
| Benchmarks Covered | 33 | 22 |
| Pricing (input/output) | $2.00 / $12.00 | $1.00 / $2.70 |

## Category Breakdown

### Agentic

- Winner: Coming soon
- GPT-5.6 Terra public-lane score: 60.7 (Supported · #17/154)
- Step 5 Preview public-lane score: Coming soon

| Benchmark | GPT-5.6 Terra | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 3.0 | 20.8% | Coming soon | Coming soon |
| Terminal-Bench 2.0 | 87.4% | Coming soon | Coming soon |
| BrowseComp | 87.5% | 88.7% | Step 5 Preview |
| OSWorld 2.0 | 50.2% | Coming soon | Coming soon |
| CyberGym | 81.8% | 84.7% | Step 5 Preview |
| ExploitGym | 23.2% | Coming soon | Coming soon |
| Toolathlon | 53.1% | Coming soon | Coming soon |
| Terminal-Bench 2.1 (Vals) | 77.5% | Coming soon | Coming soon |
| ApprenticeBench | 16% | Coming soon | Coming soon |
| Terminal-Bench 2.1 | Coming soon | 85.0% | Coming soon |
| Terminal-Bench 4.0 | Coming soon | 33.30% | Coming soon |
| AutomationBench | Coming soon | 44.0% | Coming soon |
| Toolathlon-Verified | Coming soon | 74.1% | 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 |
| DRACO | Coming soon | 83.3% | Coming soon |
| Agents' Last Exam | Coming soon | 29.5% | Coming soon |

### Coding

- Winner: Coming soon
- GPT-5.6 Terra public-lane score: 68.1 (Supported · #7/156)
- Step 5 Preview public-lane score: Coming soon

| Benchmark | GPT-5.6 Terra | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Pro | 63.4% | Coming soon | Coming soon |
| Terminal-Bench 2.0 | 87.4% | Coming soon | Coming soon |
| DeepSWE | 69.6% | 67.7% | GPT-5.6 Terra |
| FrontierCode 1.1 Extended | 55.8% | Coming soon | Coming soon |
| cursorBench32 | 64.9% | Coming soon | Coming soon |
| VulcanBench v3 | 87.0% | Coming soon | Coming soon |
| LiveCodeBench (Vals) | 85.9% | Coming soon | Coming soon |
| SWE-bench (Vals) | 95.4% | Coming soon | Coming soon |
| cursorBench40 | 41.3% | Coming soon | Coming soon |
| Terminal-Bench 2.1 | Coming soon | 85.0% | Coming soon |
| SciCode | Coming soon | 58.9% | Coming soon |
| ProgramBench | Coming soon | 80.5% | Coming soon |
| sweMarathon | Coming soon | 72.7% | Coming soon |
| MLS-Bench Lite | Coming soon | 40.5% | Coming soon |

### Reasoning

- Winner: Coming soon
- GPT-5.6 Terra public-lane score: 63.7 (#13/19)
- Step 5 Preview public-lane score: 78.5 (Unranked · 2 rankable rows)

| Benchmark | GPT-5.6 Terra | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| ARC-AGI-2 | 83.9% | Coming soon | Coming soon |
| ARC-AGI-3 | 0.8% | Coming soon | Coming soon |
| CritPt | Coming soon | 20.9% | Coming soon |

### Multimodal & Grounded

- Winner: Coming soon
- GPT-5.6 Terra public-lane score: 77.2 (#15/49)
- Step 5 Preview public-lane score: 61.2 (#28/49)

| Benchmark | GPT-5.6 Terra | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| MMMU-Pro | 80.7% | 76% | GPT-5.6 Terra |
| MMMU-Pro w/ Python | 82% | Coming soon | Coming soon |
| OfficeQA Pro | Coming soon | 60.3% | Coming soon |

### Knowledge

- Winner: Coming soon
- GPT-5.6 Terra public-lane score: 69.7 (Supported · #15/186)
- Step 5 Preview public-lane score: Coming soon

| Benchmark | GPT-5.6 Terra | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 92.9% | Coming soon | Coming soon |
| GPQA-D | 92.9% | 93.5% | Step 5 Preview |
| HLE-Verified | 51.1% | Coming soon | Coming soon |
| LABBench2 | 81.2% | Coming soon | Coming soon |
| HealthBench Professional | 57.7% | Coming soon | Coming soon |
| HealthBench Hard | 32.7% | Coming soon | Coming soon |
| GPQA Diamond (Vals) | 90.9% | Coming soon | Coming soon |
| MMLU-Pro (Vals) | 86.7% | Coming soon | Coming soon |
| HLE | Coming soon | 46.5% | Coming soon |

### Instruction Following

- Winner: Coming soon
- GPT-5.6 Terra public-lane score: 85.7 (#35/124)
- Step 5 Preview public-lane score: Coming soon

### Mathematics

- Winner: Coming soon
- GPT-5.6 Terra public-lane score: 96.8 (Unranked · 3 rankable rows)
- Step 5 Preview public-lane score: Coming soon

| Benchmark | GPT-5.6 Terra | Step 5 Preview | Winner |
|-----------|-----------|-----------|--------|
| FrontierMath (legacy) | 84.9% | Coming soon | Coming soon |
| FrontierMath v2 (Tiers 1-3) | 84.900% | Coming soon | Coming soon |
| FrontierMath v2 (Tier 4) | 68.300% | Coming soon | Coming soon |

## FAQ

### Which is better overall, GPT-5.6 Terra or Step 5 Preview?

Step 5 Preview is ahead overall on BenchLM right now.

### Where is the biggest gap between GPT-5.6 Terra and Step 5 Preview?

The widest category gap is in multimodal & grounded, where the averages are 77.2 for GPT-5.6 Terra and 61.2 for Step 5 Preview.

### How many benchmarks does GPT-5.6 Terra cover on BenchLM?

GPT-5.6 Terra currently has 33 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

- [GPT-5.6 Terra vs GPT-5.6 Sol](/compare/gpt-5-6-sol-vs-gpt-5-6-terra)
- [Step 5 Preview vs GPT-5.6 Sol](/compare/gpt-5-6-sol-vs-step-5-preview)
- [GPT-5.6 Terra vs GPT-5.6 Luna](/compare/gpt-5-6-luna-vs-gpt-5-6-terra)
- [Step 5 Preview vs GPT-5.6 Luna](/compare/gpt-5-6-luna-vs-step-5-preview)

## Explore More

- [GPT-5.6 Terra profile](/models/gpt-5-6-terra)
- [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)