Model comparison
GPT-5.4 vs Grok 4
GPT-5.4 is the stronger model overall, scoring 59.4 to 48.1 on the Noometry Index.
Last verified . 33 shared benchmarks.
Summary
- They share 33 benchmarks with published results for both. GPT-5.4 scores higher in 8 categories and Grok 4 in 2 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 leads 73.5 to 48.4.
- The biggest single-benchmark swing is ARC-AGI-2: 74% for GPT-5.4 and 16% for Grok 4.
Side by side
| GPT-5.4 | Grok 4 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 59.4 | 48.1 |
| Released | 2026-03-05 | 2025-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | — |
| Max output | 128K | — |
| Input $ / M tokens | $2.50 | — |
| Output $ / M tokens | $15 | — |
| Results tracked | 68 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 leads
GPT-5.4: 52.6 (#33), Grok 4: 50.3 (#46)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| WeirdML | 77.7% | 45.7% |
| LMArena Coding | 1497 | 1408 |
| SWE-bench Verified | 76.9% | — |
| DeepSWE | 51.8% | — |
| Aider Polyglot | — | 79.6% |
| LMArena WebDev | 1465 | — |
| SciCode | 56.6% | — |
| GSO | 31.4% | — |
| MirrorCode | 15.6% | — |
| ALE-Bench | 1,607 | — |
| AlgoTune | 1.85 | — |
Agentic & Tool Use GPT-5.4 leads
GPT-5.4: 46.5 (#13), Grok 4: 32.3 (#68)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| Terminal-Bench | 81.8% | 27.2% |
| DeepResearch Bench | 35.1% | 47.3% |
| LMArena Search | 1197 | 1142 |
| METR Time Horizons | 74.3% | 66.6% |
| APEX-Agents | 52.4% | — |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| τ²-bench Banking | 39.4% | — |
| Cybench | — | 43% |
| PostTrainBench | 19% | — |
| BALROG | — | 43.6% |
| GBAEval | 45.1% | — |
| Vending-Bench 2 | 6,144 | — |
Reasoning GPT-5.4 leads
GPT-5.4: 61.8 (#19), Grok 4: 36.7 (#65)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| ARC-AGI-2 | 74% | 16% |
| Kagi LLM Benchmark | 63.8% | 73.6% |
| ARC-AGI-1 | 93.7% | 66.7% |
| Chess Puzzles | 44% | 28% |
| LMArena Hard Prompts | 1485 | 1409 |
| Epoch Capabilities Index | 156.81 | 146.44 |
| ForecastBench | 59.5 | 60.9 |
| SimpleBench | — | 60.5% |
| NYT Connections (extended) | 91.3% | — |
| CritPt | 23.4% | — |
| EnigmaEval | 16% | — |
| Thematic Generalization | 80% | — |
| EBR-Bench | 25.4% | — |
| Mystery Game Puzzles | 37% | — |
| DTBench | 94.4% | — |
| LMCA | 52% | — |
Math GPT-5.4 leads
GPT-5.4: 73.5 (#19), Grok 4: 48.4 (#64)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 97.8% | 84% |
| LMArena Math | 1488 | 1422 |
| FrontierMath (Feb 2025 set) | 47.6% | 19.7% |
| FrontierMath Tier 4 (v1) | 27.1% | 2.1% |
| FrontierMath (Tiers 1-3) | 78.6% | — |
| FrontierMath Tier 4 | 49% | — |
| MathArena Final-Answer Competitions | 83.1% | — |
| ProofBench | 56% | — |
| Omni-MATH | — | 60.3% |
Knowledge GPT-5.4 leads
GPT-5.4: 65.3 (#14), Grok 4: 53.8 (#55)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| GPQA Diamond | 93.3% | 87% |
| LMArena Expert | 1507 | 1415 |
| Humanity's Last Exam | 36.2% | — |
| SimpleQA Verified | 45.1% | — |
| MMLU-Pro | — | 85.1% |
| Confabulations | — | 12.4% |
| Vectara Hallucination Rate | 7% | — |
| GPQA (HELM) | — | 72.7% |
Multimodal GPT-5.4 leads
GPT-5.4: 43.7 (#20), Grok 4: 33.7 (#94)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| LMArena Vision | 1303 | 1210 |
| GeoBench | — | 45% |
| Blueprint-Bench 2 | 27.1% | — |
| Furniture Assembly | 37.5% | — |
| LMArena Document | 1471 | — |
Multilingual GPT-5.4 leads
GPT-5.4: 56.2 (#23), Grok 4: 51.8 (#103)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| LMArena Non-English | 1465 | 1403 |
| LMArena Chinese | 1519 | 1427 |
| LMArena French | 1493 | 1418 |
| LMArena German | 1472 | 1429 |
| LMArena Japanese | 1485 | 1394 |
| LMArena Korean | 1448 | 1377 |
| LMArena Russian | 1480 | 1410 |
| LMArena Spanish | 1454 | 1420 |
Instruction Following Grok 4 leads
GPT-5.4: 77.1 (#27), Grok 4: 79.2 (#5)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| LMArena Instruction Following | 1469 | 1387 |
| IFEval | — | 94.9% |
Long Context Grok 4 leads
GPT-5.4: 50.3 (#8), Grok 4: 63.1 (#4)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| LMArena Longer Query | 1473 | 1409 |
| Fiction.LiveBench | — | 94.4% |
| CL-bench | 27.9% | — |
| CL-bench Life | 21.7% | — |
Writing & Preference GPT-5.4 leads
GPT-5.4: 71.9 (#17), Grok 4: 58.5 (#116)
| Benchmark | GPT-5.4 | Grok 4 |
|---|---|---|
| LMArena Text | 1469 | 1411 |
| LMArena Creative Writing | 1439 | 1397 |
| LMArena Multi-Turn | 1482 | 1416 |
| Short-Story Creative Writing | — | 76.9% |
| EQ-Bench Creative Writing | 1840 | — |
| WildBench | — | 79.7% |
| EQ-Bench 4 | 1272 | — |
Frequently asked questions
Is GPT-5.4 better than Grok 4?
GPT-5.4 is the stronger model overall, scoring 59.4 to 48.1 on the Noometry Index.
Is GPT-5.4 or Grok 4 better for coding?
GPT-5.4 scores higher on coding benchmarks: 52.6 versus 50.3 in the Noometry coding category.
How many benchmarks do GPT-5.4 and Grok 4 share?
33 benchmarks have published results for both models. GPT-5.4 has 68 scored results on Noometry and Grok 4 has 48.