# GPT-5.6 Luna vs GPT-5.6 Sol

> Side-by-side benchmark comparison for GPT-5.6 Luna and GPT-5.6 Sol across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/gpt-5-6-luna-vs-gpt-5-6-sol
- Last updated: September 24, 2026

- Shared sourced benchmarks: 29
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.7

## Quick Verdict

Pick GPT-5.6 Sol if you want the stronger benchmark profile. GPT-5.6 Luna only makes more sense when its price, context window, or workload-specific category wins matter more than the overall score.

## Summary

- GPT-5.6 Sol leads overall 78.49 to 65.6.
- The clearest category separation is in multimodal & grounded, where the averages are 67.1 for GPT-5.6 Luna and 87.6 for GPT-5.6 Sol.
- The biggest single benchmark swing is ARC-AGI-2 in Reasoning, with scores of 59.5% and 92.5%.
- GPT-5.6 Luna is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.

## Model Snapshot

| Property | GPT-5.6 Luna | GPT-5.6 Sol |
|----------|----------|----------|
| Creator | OpenAI | OpenAI |
| Type | Proprietary | Proprietary |
| Reasoning | Reasoning | Reasoning |
| Context | 1.05M | 1.05M |
| Overall Score | 65.6 | 78.49 |
| Benchmarks Covered | 29 | 37 |
| Pricing (input/output) | $0.20 / $1.20 | $4.00 / $20.00 |

## Category Breakdown

### Agentic

- Winner: GPT-5.6 Sol
- GPT-5.6 Luna public-lane score: 55.3 (Supported · #28/105)
- GPT-5.6 Sol public-lane score: 69.6 (Supported · #7/105)

| Benchmark | GPT-5.6 Luna | GPT-5.6 Sol | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 3.0 | 14.3% | 34.6% | GPT-5.6 Sol |
| Terminal-Bench 2.1 | 84.7% | 91.9% | GPT-5.6 Sol |
| BrowseComp | 83.3% | 92.2% | GPT-5.6 Sol |
| OSWorld 2.0 | 45.6% | 62.6% | GPT-5.6 Sol |
| CyberGym | 77.9% | 84.5% | GPT-5.6 Sol |
| ExploitGym | 12.4% | 33.7% | GPT-5.6 Sol |
| Toolathlon | 53.4% | 58% | GPT-5.6 Sol |
| Terminal-Bench 2.1 (Vals) | 79.0% | 85.8% | GPT-5.6 Sol |
| ApprenticeBench | 7% | 26% | GPT-5.6 Sol |

### Coding

- Winner: GPT-5.6 Sol
- GPT-5.6 Luna public-lane score: 64.5 (Supported · #9/135)
- GPT-5.6 Sol public-lane score: 71.6 (Supported · #6/135)

| Benchmark | GPT-5.6 Luna | GPT-5.6 Sol | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Pro | 62.7% | 64.6% | GPT-5.6 Sol |
| Terminal-Bench 2.1 | 84.7% | 91.9% | GPT-5.6 Sol |
| DeepSWE | 67.2% | 72.7% | GPT-5.6 Sol |
| FrontierCode 1.1 Extended | 55.1% | 60.6% | GPT-5.6 Sol |
| cursorBench32 | 61.1% | 67.2% | GPT-5.6 Sol |
| VulcanBench v3 | 85.5% | 87.0% | GPT-5.6 Sol |
| SWE-bench (Vals) | 93.0% | 96.2% | GPT-5.6 Sol |
| Bug Hunt Bench | Coming soon | 42 fixes | Coming soon |
| FrontierSWE v2 | Coming soon | 32.2% | Coming soon |
| VulcanBench CII v1 | Coming soon | 86.5% | Coming soon |
| LiveCodeBench (Vals) | Coming soon | 82.6% | Coming soon |
| cursorBench40 | Coming soon | 41.7% | Coming soon |

### Reasoning

- Winner: GPT-5.6 Sol
- GPT-5.6 Luna public-lane score: 54.7 (#18/19)
- GPT-5.6 Sol public-lane score: 72.1 (#8/19)

| Benchmark | GPT-5.6 Luna | GPT-5.6 Sol | Winner |
|-----------|-----------|-----------|--------|
| ARC-AGI-2 | 59.5% | 92.5% | GPT-5.6 Sol |
| ARC-AGI-3 | 0.2% | 7.8% | GPT-5.6 Sol |
| GeneBench-Pro | Coming soon | 28.7% | Coming soon |

### Multimodal & Grounded

- Winner: GPT-5.6 Sol
- GPT-5.6 Luna public-lane score: 67.1 (#22/50)
- GPT-5.6 Sol public-lane score: 87.6 (#5/50)

| Benchmark | GPT-5.6 Luna | GPT-5.6 Sol | Winner |
|-----------|-----------|-----------|--------|
| MMMU-Pro | 78.4% | 83% | GPT-5.6 Sol |
| MMMU-Pro w/ Python | 79.5% | 84.6% | GPT-5.6 Sol |

### Knowledge

- Winner: GPT-5.6 Sol
- GPT-5.6 Luna public-lane score: 64.6 (Supported · #22/158)
- GPT-5.6 Sol public-lane score: 78.8 (Supported · #7/158)

| Benchmark | GPT-5.6 Luna | GPT-5.6 Sol | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 92.3% | 94.6% | GPT-5.6 Sol |
| GPQA-D | 92.3% | 94.6% | GPT-5.6 Sol |
| HealthBench Professional | 55.7% | 60.5% | GPT-5.6 Sol |
| HealthBench Hard | 32.0% | 33.1% | GPT-5.6 Sol |
| GPQA Diamond (Vals) | 91.7% | 95.2% | GPT-5.6 Sol |
| MMLU-Pro (Vals) | 86.0% | 89.1% | GPT-5.6 Sol |
| HLE-Verified | Coming soon | 54.5% | Coming soon |
| LABBench2 | Coming soon | 82.1% | Coming soon |

### Instruction Following

- Winner: Not comparable
- GPT-5.6 Luna public-lane score: Coming soon
- GPT-5.6 Sol public-lane score: 87.7 (#28/124)

### Mathematics

- Winner: Not comparable
- GPT-5.6 Luna public-lane score: 94.2 (Unranked · 3 rankable rows)
- GPT-5.6 Sol public-lane score: 96.8 (Unranked · 3 rankable rows)

| Benchmark | GPT-5.6 Luna | GPT-5.6 Sol | Winner |
|-----------|-----------|-----------|--------|
| FrontierMath (legacy) | 78.6% | 89% | GPT-5.6 Sol |
| FrontierMath v2 (Tiers 1-3) | 78.600% | 89.000% | GPT-5.6 Sol |
| FrontierMath v2 (Tier 4) | 58.500% | 83.000% | GPT-5.6 Sol |

## FAQ

### Which is better overall, GPT-5.6 Luna or GPT-5.6 Sol?

GPT-5.6 Sol is ahead overall on BenchLM right now.

### Where is the biggest gap between GPT-5.6 Luna and GPT-5.6 Sol?

The widest category gap is in multimodal & grounded, where the averages are 67.1 for GPT-5.6 Luna and 87.6 for GPT-5.6 Sol.

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

GPT-5.6 Luna currently has 29 sourced benchmark scores on BenchLM.

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

GPT-5.6 Sol currently has 37 sourced benchmark scores on BenchLM.

## Related Comparisons

- [GPT-5.6 Luna vs GPT-5.6 Terra](/compare/gpt-5-6-luna-vs-gpt-5-6-terra)
- [GPT-5.6 Sol vs GPT-5.6 Terra](/compare/gpt-5-6-sol-vs-gpt-5-6-terra)
- [GPT-5.6 Luna vs GPT-5.6 Cyber](/compare/gpt-5-6-cyber-vs-gpt-5-6-luna)
- [GPT-5.6 Sol vs GPT-5.6 Cyber](/compare/gpt-5-6-cyber-vs-gpt-5-6-sol)

## Explore More

- [GPT-5.6 Luna profile](/models/gpt-5-6-luna)
- [GPT-5.6 Sol profile](/models/gpt-5-6-sol)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)