Model comparison
GPT-5.6 Luna vs GPT-5 Mini
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 41.8 on the Noometry Index.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. GPT-5.6 Luna scores higher in 9 categories and GPT-5 Mini in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.7 to 46.7.
- The biggest single-benchmark swing is ARC-AGI-2: 59.5% for GPT-5.6 Luna and 4.4% for GPT-5 Mini.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $0.25 / $2 for GPT-5 Mini.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 400K.
Side by side
| GPT-5.6 Luna | GPT-5 Mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 54.6 | 41.8 |
| Released | 2026-07-09 | 2025-08-07 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 400K |
| Max output | 128K | 128K |
| Input $ / M tokens | $0.20 | $0.25 |
| Output $ / M tokens | $1.20 | $2 |
| Results tracked | 52 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.6 Luna leads
GPT-5.6 Luna: 54.5 (#28), GPT-5 Mini: 40.1 (#146)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| SciCode | 53.6% | 39.2% |
| WeirdML | 60.9% | 52.7% |
| LMArena Coding | 1466 | 1406 |
| ALE-Bench | 1,667 | 799.77 |
| SWE-bench Verified | — | 64.7% |
| DeepSWE | 67.2% | — |
| FrontierCode | 39.8% | — |
| SWE-bench Verified (bash only) | — | 59.8% |
| CursorBench | 35.9% | — |
| LMArena WebDev | 1519 | — |
| SWE-bench Multilingual | — | 39.7% |
| AlgoTune | — | 1.38 |
Agentic & Tool Use GPT-5.6 Luna leads
GPT-5.6 Luna: 34.4 (#45), GPT-5 Mini: 31.1 (#70)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| Vending-Bench 2 | 4,095 | -31.18 |
| Terminal-Bench | — | 34.8% |
| APEX-Agents | 43% | — |
| Berkeley Function Calling Leaderboard | — | 55.5% |
| BALROG | 45.6% | — |
| GDP.pdf | 22.7% | — |
Reasoning GPT-5.6 Luna leads
GPT-5.6 Luna: 47.6 (#43), GPT-5 Mini: 23.9 (#168)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| ARC-AGI-2 | 59.5% | 4.4% |
| Kagi LLM Benchmark | 49.1% | 70.3% |
| ARC-AGI-1 | 88% | 54.3% |
| CritPt | 20.6% | 0% |
| Chess Puzzles | 40% | 30% |
| LMArena Hard Prompts | 1451 | 1380 |
| Mystery Game Puzzles | 21% | 10% |
| DTBench | 89.1% | 80.5% |
| LMCA | 48.5% | 34.2% |
| Epoch Capabilities Index | 156.39 | 145.52 |
| SimpleBench | 46.8% | — |
| NYT Connections (extended) | 69.4% | — |
| EnigmaEval | — | 8.2% |
| Surface Evolver Bench | 61.9% | — |
| ForecastBench | — | 61 |
Math GPT-5.6 Luna leads
GPT-5.6 Luna: 77.7 (#14), GPT-5 Mini: 46.7 (#69)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 82.1% | 46.7% |
| FrontierMath Tier 4 | 61% | 12.2% |
| OTIS Mock AIME 2024-2025 | 98.3% | 86.7% |
| ProofBench | 60% | 9% |
| LMArena Math | 1458 | 1378 |
| Omni-MATH | — | 72.2% |
| MATH Level 5 | — | 97.8% |
| FrontierMath (Feb 2025 set) | — | 27.2% |
| FrontierMath Tier 4 (v1) | — | 6.3% |
Knowledge GPT-5.6 Luna leads
GPT-5.6 Luna: 58.5 (#34), GPT-5 Mini: 45.6 (#86)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| GPQA Diamond | 91.6% | 75% |
| SimpleQA Verified | 41% | 21.6% |
| LMArena Expert | 1478 | 1379 |
| Humanity's Last Exam | — | 19.4% |
| MMLU-Pro | — | 83.5% |
| Confabulations | — | 13.3% |
| Vectara Hallucination Rate | — | 12.9% |
| GPQA (HELM) | — | 75.6% |
Multimodal GPT-5.6 Luna leads
GPT-5.6 Luna: 42.7 (#28), GPT-5 Mini: 35.6 (#85)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| LMArena Vision | 1258 | 1202 |
| VPCT | — | 40.2% |
| Blueprint-Bench 2 | 22.6% | — |
| Furniture Assembly | 42.5% | — |
| LMArena Document | 1457 | — |
Multilingual GPT-5.6 Luna leads
GPT-5.6 Luna: 52.8 (#78), GPT-5 Mini: 48.9 (#137)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| LMArena Non-English | 1417 | 1363 |
| LMArena Chinese | 1470 | 1385 |
| LMArena French | 1456 | 1386 |
| LMArena German | 1454 | 1366 |
| LMArena Japanese | 1411 | 1341 |
| LMArena Korean | 1415 | 1308 |
| LMArena Russian | 1428 | 1362 |
| LMArena Spanish | 1448 | 1355 |
Instruction Following Too close to call
GPT-5.6 Luna: 75.6 (#57), GPT-5 Mini: 76.2 (#46)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| LMArena Instruction Following | 1437 | 1357 |
| IFEval | — | 92.7% |
Long Context GPT-5.6 Luna leads
GPT-5.6 Luna: 43.9 (#82), GPT-5 Mini: 41.9 (#132)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| LMArena Longer Query | 1436 | 1355 |
| Fiction.LiveBench | — | 69.4% |
Writing & Preference GPT-5.6 Luna leads
GPT-5.6 Luna: 68.0 (#29), GPT-5 Mini: 55.2 (#148)
| Benchmark | GPT-5.6 Luna | GPT-5 Mini |
|---|---|---|
| LMArena Text | 1431 | 1373 |
| LMArena Creative Writing | 1396 | 1325 |
| EQ-Bench Creative Writing | 1829 | 1313 |
| LMArena Multi-Turn | 1434 | 1363 |
| Short-Story Creative Writing | — | 83.1% |
| WildBench | — | 85.5% |
| EQ-Bench 4 | 1156 | — |
Frequently asked questions
Is GPT-5.6 Luna better than GPT-5 Mini?
GPT-5.6 Luna is the stronger model overall, scoring 54.6 to 41.8 on the Noometry Index.
Which is cheaper, GPT-5.6 Luna or GPT-5 Mini?
GPT-5.6 Luna is cheaper. It lists at $0.20 per million input tokens and $1.20 per million output tokens; GPT-5 Mini lists at $0.25 and $2.
Is GPT-5.6 Luna or GPT-5 Mini better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 54.5 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 400K.
How many benchmarks do GPT-5.6 Luna and GPT-5 Mini share?
38 benchmarks have published results for both models. GPT-5.6 Luna has 52 scored results on Noometry and GPT-5 Mini has 60.