Model comparison
GPT-5.6 Luna vs Grok 3
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 39.6 on the Noometry Index.
Last verified . 27 shared benchmarks.
Summary
- They share 27 benchmarks with published results for both. GPT-5.6 Luna scores higher in 9 categories and Grok 3 in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Luna leads 77.5 to 36.2.
- The biggest single-benchmark swing is ARC-AGI-1: 88% for GPT-5.6 Luna and 5.5% for Grok 3.
Side by side
| GPT-5.6 Luna | Grok 3 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 54.7 | 39.6 |
| Released | 2026-07-09 | 2025-04-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.20 | — |
| Output $ / M tokens | $1.20 | — |
| Results tracked | 51 | 40 |
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: 53.9 (#25), Grok 3: 41.5 (#113)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| WeirdML | 60.9% | 37.2% |
| LMArena Coding | 1466 | 1432 |
| DeepSWE | 67.2% | — |
| FrontierCode | 39.8% | — |
| Aider Polyglot | — | 53.3% |
| CursorBench | 35.9% | — |
| LMArena WebDev | 1517 | — |
| SciCode | 53.6% | — |
| ALE-Bench | 1,667 | — |
Agentic & Tool Use GPT-5.6 Luna leads
GPT-5.6 Luna: 35.3 (#37), Grok 3: 30.2 (#69)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| BALROG | 45.6% | 29.5% |
| GDP.pdf | 22.7% | — |
| Vending-Bench 2 | 4,095 | — |
Reasoning GPT-5.6 Luna leads
GPT-5.6 Luna: 47.5 (#38), Grok 3: 13.8 (#322)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| ARC-AGI-2 | 59.5% | 0% |
| SimpleBench | 46.8% | 36.1% |
| Kagi LLM Benchmark | 49.1% | 61.3% |
| ARC-AGI-1 | 88% | 5.5% |
| LMArena Hard Prompts | 1452 | 1433 |
| Epoch Capabilities Index | 156.32 | 138.34 |
| NYT Connections (extended) | 69.4% | — |
| CritPt | 20.6% | — |
| Chess Puzzles | 40% | — |
| Mystery Game Puzzles | 21% | — |
| DTBench | 89.1% | — |
| LMCA | 48.5% | — |
| Surface Evolver Bench | 61.9% | — |
Math GPT-5.6 Luna leads
GPT-5.6 Luna: 77.5 (#12), Grok 3: 36.2 (#122)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.3% | 55.6% |
| LMArena Math | 1455 | 1390 |
| FrontierMath (Tiers 1-3) | 82.1% | — |
| FrontierMath Tier 4 | 61% | — |
| ProofBench | 60% | — |
| Omni-MATH | — | 46.4% |
| MATH Level 5 | — | 88.7% |
| FrontierMath (Feb 2025 set) | — | 3.8% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge GPT-5.6 Luna leads
GPT-5.6 Luna: 58.6 (#31), Grok 3: 46.0 (#76)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| GPQA Diamond | 91.6% | 75.8% |
| LMArena Expert | 1479 | 1420 |
| SimpleQA Verified | 41% | — |
| MMLU-Pro | — | 78.8% |
| Confabulations | — | 14.2% |
| Vectara Hallucination Rate | — | 5.8% |
| GPQA (HELM) | — | 65% |
Multimodal Not comparable
GPT-5.6 Luna: 43.0 (#22), Grok 3: —
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| LMArena Vision | 1259 | — |
| Blueprint-Bench 2 | 22.6% | — |
| Furniture Assembly | 42.5% | — |
| LMArena Document | 1457 | — |
Multilingual Too close to call
GPT-5.6 Luna: 53.0 (#73), Grok 3: 52.5 (#81)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| LMArena Non-English | 1418 | 1411 |
| LMArena Chinese | 1469 | 1448 |
| LMArena French | 1457 | 1460 |
| LMArena German | 1459 | 1431 |
| LMArena Japanese | 1416 | 1389 |
| LMArena Korean | 1414 | 1374 |
| LMArena Russian | 1428 | 1416 |
| LMArena Spanish | 1445 | 1417 |
Instruction Following Too close to call
GPT-5.6 Luna: 76.0 (#48), Grok 3: 75.2 (#69)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| LMArena Instruction Following | 1440 | 1409 |
| IFEval | — | 88.4% |
Long Context GPT-5.6 Luna leads
GPT-5.6 Luna: 43.9 (#78), Grok 3: 38.7 (#187)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| LMArena Longer Query | 1437 | 1439 |
| Fiction.LiveBench | — | 58.3% |
Writing & Preference GPT-5.6 Luna leads
GPT-5.6 Luna: 68.3 (#27), Grok 3: 56.0 (#136)
| Benchmark | GPT-5.6 Luna | Grok 3 |
|---|---|---|
| LMArena Text | 1432 | 1426 |
| LMArena Creative Writing | 1397 | 1414 |
| EQ-Bench Creative Writing | 1829 | 1186 |
| LMArena Multi-Turn | 1433 | 1424 |
| Short-Story Creative Writing | — | 76.4% |
| WildBench | — | 84.9% |
| EQ-Bench 4 | 1156 | — |
Frequently asked questions
Is GPT-5.6 Luna better than Grok 3?
GPT-5.6 Luna is the stronger model overall, scoring 54.7 to 39.6 on the Noometry Index.
Is GPT-5.6 Luna or Grok 3 better for coding?
GPT-5.6 Luna scores higher on coding benchmarks: 53.9 versus 41.5 in the Noometry coding category.
How many benchmarks do GPT-5.6 Luna and Grok 3 share?
27 benchmarks have published results for both models. GPT-5.6 Luna has 51 scored results on Noometry and Grok 3 has 40.