Model comparison
GPT-6 Luna vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 53.3 on the Noometry Index. GPT-6 Luna costs 15× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Last verified . 41 shared benchmarks.
Summary
- They share 41 benchmarks with published results for both. GPT-6 Luna scores higher in 3 categories and Grok 4.5 in 7 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Luna leads 76.1 to 60.9.
- The biggest single-benchmark swing is ProofBench: 64% for GPT-6 Luna and 31% for Grok 4.5.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $2 / $6 for Grok 4.5.
- GPT-6 Luna accepts more context: 1.05M tokens versus 500K.
Side by side
| GPT-6 Luna | Grok 4.5 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 53.3 | 55.0 |
| Released | 2026-09-22 | 2026-07-08 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $0.10 | $2 |
| Output $ / M tokens | $0.50 | $6 |
| Results tracked | 42 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Luna leads
GPT-6 Luna: 55.5 (#25), Grok 4.5: 52.2 (#35)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| DeepSWE | 66.6% | 53.8% |
| FrontierCode | 42.4% | 42.4% |
| LMArena WebDev | 1581 | 1553 |
| SciCode | 54.6% | 54.1% |
| LMArena Coding | 1439 | 1474 |
| ALE-Bench | 1,577 | 1,309 |
| WeirdML | — | 46.4% |
Agentic & Tool Use Grok 4.5 leads
GPT-6 Luna: 33.3 (#54), Grok 4.5: 44.4 (#17)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| APEX-Agents | 44.3% | 56.2% |
| GDP.pdf | 23% | 14% |
| τ²-bench Banking | — | 47.9% |
| PostTrainBench | — | 23.4% |
| GBAEval | — | 65.4% |
| LMArena Search | — | 1213 |
| Vending-Bench 2 | — | 3,887 |
Reasoning Grok 4.5 leads
GPT-6 Luna: 48.2 (#41), Grok 4.5: 56.1 (#25)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| ARC-AGI-2 | 59.3% | 52.6% |
| NYT Connections (extended) | 68.7% | 79.9% |
| ARC-AGI-1 | 86.7% | 87.2% |
| CritPt | 19.4% | 15.4% |
| Chess Puzzles | 31% | 36% |
| LMArena Hard Prompts | 1411 | 1462 |
| DTBench | 90.1% | 96.5% |
| LMCA | 44.5% | 45.2% |
| Epoch Capabilities Index | 156.28 | 153.92 |
| SimpleBench | — | 70% |
| Kagi LLM Benchmark | — | 83.5% |
| Mystery Game Puzzles | 7% | — |
| Surface Evolver Bench | — | 74.4% |
Math GPT-6 Luna leads
GPT-6 Luna: 76.1 (#15), Grok 4.5: 60.9 (#35)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 78.9% | 57.2% |
| FrontierMath Tier 4 | 56.1% | 24.4% |
| OTIS Mock AIME 2024-2025 | 98.9% | 97.8% |
| ProofBench | 64% | 31% |
| LMArena Math | 1416 | 1459 |
Knowledge Grok 4.5 leads
GPT-6 Luna: 57.0 (#41), Grok 4.5: 62.3 (#24)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 90.5% | 93.4% |
| SimpleQA Verified | 41.4% | 48.3% |
| LMArena Expert | 1444 | 1466 |
Multimodal GPT-6 Luna leads
GPT-6 Luna: 42.4 (#30), Grok 4.5: 37.6 (#72)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| LMArena Vision | 1217 | 1288 |
| Blueprint-Bench 2 | 31.2% | 27.3% |
| Furniture Assembly | 44.2% | 22.5% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.5 leads
GPT-6 Luna: 50.5 (#117), Grok 4.5: 54.4 (#42)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1386 | 1440 |
| LMArena Chinese | 1433 | 1496 |
| LMArena French | 1420 | 1456 |
| LMArena German | 1369 | 1446 |
| LMArena Japanese | 1369 | 1428 |
| LMArena Korean | 1360 | 1404 |
| LMArena Russian | 1394 | 1448 |
| LMArena Spanish | 1393 | 1450 |
Instruction Following Grok 4.5 leads
GPT-6 Luna: 74.3 (#99), Grok 4.5: 76.0 (#48)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1409 | 1446 |
Long Context Grok 4.5 leads
GPT-6 Luna: 43.0 (#111), Grok 4.5: 44.8 (#56)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1409 | 1463 |
Writing & Preference Grok 4.5 leads
GPT-6 Luna: 58.3 (#119), Grok 4.5: 65.8 (#42)
| Benchmark | GPT-6 Luna | Grok 4.5 |
|---|---|---|
| LMArena Text | 1391 | 1448 |
| LMArena Creative Writing | 1363 | 1442 |
| LMArena Multi-Turn | 1396 | 1456 |
| EQ-Bench Creative Writing | — | 1579 |
Frequently asked questions
Is GPT-6 Luna better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 53.3 on the Noometry Index. GPT-6 Luna costs 15× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Which is cheaper, GPT-6 Luna or Grok 4.5?
GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Grok 4.5 lists at $2 and $6.
Is GPT-6 Luna or Grok 4.5 better for coding?
GPT-6 Luna scores higher on coding benchmarks: 55.5 versus 52.2 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Luna does, with 1.05M tokens against 500K.
How many benchmarks do GPT-6 Luna and Grok 4.5 share?
41 benchmarks have published results for both models. GPT-6 Luna has 42 scored results on Noometry and Grok 4.5 has 52.