# GPT-5.6 Sol vs GPT-5 Mini

> GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 41.7 on the Noometry Index. GPT-5 Mini costs 12× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-6-sol-vs-gpt-5-mini
- Last updated: 2026-10-10
- Shared benchmarks: 40

## Summary

- They share 40 benchmarks with published results for both. GPT-5.6 Sol scores higher in 10 categories and GPT-5 Mini in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.6 Sol leads 74.7 to 24.1.
- The biggest single-benchmark swing is ARC-AGI-2: 92.5% for GPT-5.6 Sol and 4.4% for GPT-5 Mini.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $4 / $20 for GPT-5.6 Sol.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 400K.

## Snapshot

| | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 64.4 | 41.7 |
| Rank | 6 | 115 |
| Context | 1.05M | 400K |
| Input $/M | $4 | $0.25 |
| Output $/M | $20 | $2 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-5.6 Sol: 61.1 (#7)
- GPT-5 Mini: 40.0 (#134)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| SciCode | 57.1% | 39.2% |
| WeirdML | 89.4% | 52.7% |
| LMArena Coding | 1493 | 1406 |
| ALE-Bench | 2,177 | 799.77 |
| SWE-bench Verified | — | 64.7% |
| DeepSWE | 72.7% | — |
| FrontierCode | 47.5% | — |
| SWE-bench Verified (bash only) | — | 59.8% |
| CursorBench | 41.7% | — |
| LMArena WebDev | 1619 | — |
| SWE-bench Multilingual | — | 39.7% |
| FrontierSWE | 32.2% | — |
| MirrorCode | 20% | — |
| AlgoTune | — | 1.38 |

## Agentic & Tool Use

- GPT-5.6 Sol: 50.2 (#7)
- GPT-5 Mini: 31.0 (#64)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| Vending-Bench 2 | 9,619 | -31.18 |
| Terminal-Bench | — | 34.8% |
| APEX-Agents | 51.4% | — |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| OSWorld 2.0 | 27.3% | — |
| τ²-bench Banking | 46.9% | — |
| PostTrainBench | 36.2% | — |
| BALROG | 60% | — |
| GBAEval | 52.6% | — |
| GDP.pdf | 30.7% | — |
| LMArena Search | 1257 | — |

## Reasoning

- GPT-5.6 Sol: 74.7 (#6)
- GPT-5 Mini: 24.1 (#153)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| ARC-AGI-2 | 92.5% | 4.4% |
| Kagi LLM Benchmark | 67% | 70.3% |
| ARC-AGI-1 | 97.5% | 54.3% |
| CritPt | 32.3% | 0% |
| Chess Puzzles | 64% | 30% |
| EnigmaEval | 37.1% | 8.2% |
| LMArena Hard Prompts | 1481 | 1380 |
| Mystery Game Puzzles | 58% | 10% |
| DTBench | 96% | 80.5% |
| LMCA | 59.2% | 34.2% |
| Epoch Capabilities Index | 161.99 | 145.51 |
| SimpleBench | 71.7% | — |
| NYT Connections (extended) | 93.8% | — |
| EBR-Bench | 44.8% | — |
| Surface Evolver Bench | 93.1% | — |
| Bench to the Future 3 | 0.14 | — |
| ForecastBench | — | 61 |

## Math

- GPT-5.6 Sol: 85.5 (#7)
- GPT-5 Mini: 46.3 (#59)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 89.1% | 46.7% |
| FrontierMath Tier 4 | 82.9% | 12.2% |
| OTIS Mock AIME 2024-2025 | 100% | 86.7% |
| ProofBench | 83% | 9% |
| LMArena Math | 1472 | 1377 |
| Omni-MATH | — | 72.2% |
| MATH Level 5 | — | 97.8% |
| FrontierMath (Feb 2025 set) | — | 27.2% |
| FrontierMath Erdős | 0% | — |
| FrontierMath Tier 4 (v1) | — | 6.3% |

## Knowledge

- GPT-5.6 Sol: 64.3 (#16)
- GPT-5 Mini: 45.5 (#80)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| GPQA Diamond | 93.5% | 75% |
| SimpleQA Verified | 69.7% | 21.6% |
| Vectara Hallucination Rate | 12.4% | 12.9% |
| LMArena Expert | 1517 | 1377 |
| Humanity's Last Exam | — | 19.4% |
| MMLU-Pro | — | 83.5% |
| Confabulations | — | 13.3% |
| GPQA (HELM) | — | 75.6% |

## Multimodal

- GPT-5.6 Sol: 48.9 (#8)
- GPT-5 Mini: 35.0 (#80)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| LMArena Vision | 1281 | 1202 |
| VPCT | — | 40.2% |
| Blueprint-Bench 2 | 33.6% | — |
| Furniture Assembly | 56.7% | — |
| LMArena Document | 1483 | — |

## Multilingual

- GPT-5.6 Sol: 55.4 (#32)
- GPT-5 Mini: 49.1 (#131)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1450 | 1363 |
| LMArena Chinese | 1523 | 1387 |
| LMArena French | 1475 | 1385 |
| LMArena German | 1476 | 1366 |
| LMArena Japanese | 1475 | 1342 |
| LMArena Korean | 1434 | 1306 |
| LMArena Russian | 1466 | 1362 |
| LMArena Spanish | 1449 | 1355 |

## Instruction Following

- GPT-5.6 Sol: 77.8 (#16)
- GPT-5 Mini: 76.4 (#43)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| LMArena Instruction Following | 1480 | 1357 |
| IFEval | — | 92.7% |

## Long Context

- GPT-5.6 Sol: 45.3 (#42)
- GPT-5 Mini: 41.9 (#127)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| LMArena Longer Query | 1479 | 1355 |
| Fiction.LiveBench | — | 69.4% |

## Writing & Preference

- GPT-5.6 Sol: 73.5 (#12)
- GPT-5 Mini: 55.4 (#143)

| Benchmark | GPT-5.6 Sol | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1456 | 1373 |
| LMArena Creative Writing | 1450 | 1325 |
| EQ-Bench Creative Writing | 1972 | 1313 |
| LMArena Multi-Turn | 1456 | 1363 |
| Short-Story Creative Writing | — | 83.1% |
| WildBench | — | 85.5% |
| EQ-Bench 4 | 1250 | — |

## FAQ

### Is GPT-5.6 Sol better than GPT-5 Mini?

GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 41.7 on the Noometry Index. GPT-5 Mini costs 12× less per token, which makes it the better buy when GPT-5.6 Sol's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.6 Sol or GPT-5 Mini?

GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; GPT-5.6 Sol lists at $4 and $20.

### Is GPT-5.6 Sol or GPT-5 Mini better for coding?

GPT-5.6 Sol scores higher on coding benchmarks: 61.1 versus 40.0 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.6 Sol does, with 1.05M tokens against 400K.

### How many benchmarks do GPT-5.6 Sol and GPT-5 Mini share?

40 benchmarks have published results for both models. GPT-5.6 Sol has 64 scored results on Noometry and GPT-5 Mini has 60.