Model comparison
GPT-5.6 Luna vs Grok 4.6
Grok 4.6 is the stronger model overall, scoring 56.9 to 54.6 on the Noometry Index. GPT-5.6 Luna costs 6.7× less per token, which makes it the better buy when Grok 4.6's lead doesn't matter for your workload.
Last verified . 47 shared benchmarks.
Summary
- They share 47 benchmarks with published results for both. GPT-5.6 Luna scores higher in 3 categories and Grok 4.6 in 7 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.6 leads 61.4 to 47.6.
- The biggest single-benchmark swing is FrontierMath Tier 4: 61% for GPT-5.6 Luna and 31.7% for Grok 4.6.
- GPT-5.6 Luna is cheaper at $0.20 / $1.20 per million input/output tokens, against $2 / $6 for Grok 4.6.
- GPT-5.6 Luna accepts more context: 1.05M tokens versus 500K.
Side by side
| GPT-5.6 Luna | Grok 4.6 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 54.6 | 56.9 |
| Released | 2026-07-09 | 2026-08-12 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $0.20 | $2 |
| Output $ / M tokens | $1.20 | $6 |
| Results tracked | 52 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.6 leads
GPT-5.6 Luna: 54.5 (#28), Grok 4.6: 58.5 (#16)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| DeepSWE | 67.2% | 67.5% |
| FrontierCode | 39.8% | 48% |
| CursorBench | 35.9% | 41.4% |
| LMArena WebDev | 1519 | 1617 |
| SciCode | 53.6% | 56.5% |
| WeirdML | 60.9% | 67.3% |
| LMArena Coding | 1466 | 1465 |
| ALE-Bench | 1,667 | 1,508 |
| FrontierSWE | — | 25.3% |
Agentic & Tool Use Grok 4.6 leads
GPT-5.6 Luna: 34.4 (#45), Grok 4.6: 39.4 (#27)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| APEX-Agents | 43% | 65.3% |
| GDP.pdf | 22.7% | 17.2% |
| Vending-Bench 2 | 4,095 | 9,047 |
| BALROG | 45.6% | — |
Reasoning Grok 4.6 leads
GPT-5.6 Luna: 47.6 (#43), Grok 4.6: 61.4 (#20)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| ARC-AGI-2 | 59.5% | 67.1% |
| SimpleBench | 46.8% | 75.9% |
| NYT Connections (extended) | 69.4% | 80% |
| ARC-AGI-1 | 88% | 87.5% |
| CritPt | 20.6% | 19.7% |
| Chess Puzzles | 40% | 40% |
| LMArena Hard Prompts | 1451 | 1447 |
| Mystery Game Puzzles | 21% | 34% |
| DTBench | 89.1% | 97.3% |
| LMCA | 48.5% | 48.5% |
| Epoch Capabilities Index | 156.39 | 156.44 |
| Kagi LLM Benchmark | 49.1% | — |
| EBR-Bench | — | 30.5% |
| Surface Evolver Bench | 61.9% | — |
Math GPT-5.6 Luna leads
GPT-5.6 Luna: 77.7 (#14), Grok 4.6: 67.0 (#24)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 82.1% | 66% |
| FrontierMath Tier 4 | 61% | 31.7% |
| OTIS Mock AIME 2024-2025 | 98.3% | 99.2% |
| ProofBench | 60% | 51% |
| LMArena Math | 1458 | 1423 |
Knowledge Grok 4.6 leads
GPT-5.6 Luna: 58.5 (#34), Grok 4.6: 63.3 (#20)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 91.6% | 94% |
| SimpleQA Verified | 41% | 49.3% |
| LMArena Expert | 1478 | 1467 |
Multimodal Too close to call
GPT-5.6 Luna: 42.7 (#28), Grok 4.6: 43.6 (#23)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| LMArena Vision | 1258 | 1263 |
| Blueprint-Bench 2 | 22.6% | 33.2% |
| Furniture Assembly | 42.5% | 40% |
| LMArena Document | 1457 | 1452 |
Multilingual Too close to call
GPT-5.6 Luna: 52.8 (#78), Grok 4.6: 53.0 (#74)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1417 | 1420 |
| LMArena Chinese | 1470 | 1480 |
| LMArena French | 1456 | 1461 |
| LMArena German | 1454 | 1431 |
| LMArena Japanese | 1411 | 1376 |
| LMArena Korean | 1415 | 1397 |
| LMArena Russian | 1428 | 1422 |
| LMArena Spanish | 1448 | 1404 |
Instruction Following Too close to call
GPT-5.6 Luna: 75.6 (#57), Grok 4.6: 75.4 (#63)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1437 | 1431 |
Long Context Too close to call
GPT-5.6 Luna: 43.9 (#82), Grok 4.6: 44.5 (#66)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1436 | 1454 |
Writing & Preference GPT-5.6 Luna leads
GPT-5.6 Luna: 68.0 (#29), Grok 4.6: 62.3 (#80)
| Benchmark | GPT-5.6 Luna | Grok 4.6 |
|---|---|---|
| LMArena Text | 1431 | 1428 |
| LMArena Creative Writing | 1396 | 1428 |
| LMArena Multi-Turn | 1434 | 1425 |
| EQ-Bench Creative Writing | 1829 | — |
| EQ-Bench 4 | 1156 | — |
Frequently asked questions
Is GPT-5.6 Luna better than Grok 4.6?
Grok 4.6 is the stronger model overall, scoring 56.9 to 54.6 on the Noometry Index. GPT-5.6 Luna costs 6.7× less per token, which makes it the better buy when Grok 4.6's lead doesn't matter for your workload.
Which is cheaper, GPT-5.6 Luna or Grok 4.6?
GPT-5.6 Luna is cheaper. It lists at $0.20 per million input tokens and $1.20 per million output tokens; Grok 4.6 lists at $2 and $6.
Is GPT-5.6 Luna or Grok 4.6 better for coding?
Grok 4.6 scores higher on coding benchmarks: 58.5 versus 54.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.6 Luna does, with 1.05M tokens against 500K.
How many benchmarks do GPT-5.6 Luna and Grok 4.6 share?
47 benchmarks have published results for both models. GPT-5.6 Luna has 52 scored results on Noometry and Grok 4.6 has 49.