Model comparison
GPT-5.4 vs Grok 4.5
GPT-5.4 is the stronger model overall, scoring 59.4 to 55.0 on the Noometry Index. Grok 4.5 costs 1.9× less per token, which makes it the better buy when GPT-5.4's lead doesn't matter for your workload.
Last verified . 48 shared benchmarks.
Summary
- They share 48 benchmarks with published results for both. GPT-5.4 scores higher in 10 categories and Grok 4.5 in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 leads 73.5 to 60.9.
- The biggest single-benchmark swing is WeirdML: 77.7% for GPT-5.4 and 46.4% for Grok 4.5.
- Grok 4.5 is cheaper at $2 / $6 per million input/output tokens, against $2.50 / $15 for GPT-5.4.
- GPT-5.4 accepts more context: 1.05M tokens versus 500K.
Side by side
| GPT-5.4 | Grok 4.5 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 59.4 | 55.0 |
| Released | 2026-03-05 | 2026-07-08 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $2.50 | $2 |
| Output $ / M tokens | $15 | $6 |
| Results tracked | 68 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-5.4: 52.6 (#33), Grok 4.5: 52.2 (#35)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| DeepSWE | 51.8% | 53.8% |
| LMArena WebDev | 1465 | 1553 |
| SciCode | 56.6% | 54.1% |
| WeirdML | 77.7% | 46.4% |
| LMArena Coding | 1497 | 1474 |
| ALE-Bench | 1,607 | 1,309 |
| SWE-bench Verified | 76.9% | — |
| FrontierCode | — | 42.4% |
| GSO | 31.4% | — |
| MirrorCode | 15.6% | — |
| AlgoTune | 1.85 | — |
Agentic & Tool Use GPT-5.4 leads
GPT-5.4: 46.5 (#13), Grok 4.5: 44.4 (#17)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| APEX-Agents | 52.4% | 56.2% |
| τ²-bench Banking | 39.4% | 47.9% |
| PostTrainBench | 19% | 23.4% |
| GBAEval | 45.1% | 65.4% |
| LMArena Search | 1197 | 1213 |
| Vending-Bench 2 | 6,144 | 3,887 |
| Terminal-Bench | 81.8% | — |
| DeepResearch Bench | 35.1% | — |
| GDP.pdf | — | 14% |
| METR Time Horizons | 74.3% | — |
Reasoning GPT-5.4 leads
GPT-5.4: 61.8 (#19), Grok 4.5: 56.1 (#25)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| ARC-AGI-2 | 74% | 52.6% |
| Kagi LLM Benchmark | 63.8% | 83.5% |
| NYT Connections (extended) | 91.3% | 79.9% |
| ARC-AGI-1 | 93.7% | 87.2% |
| CritPt | 23.4% | 15.4% |
| Chess Puzzles | 44% | 36% |
| LMArena Hard Prompts | 1485 | 1462 |
| DTBench | 94.4% | 96.5% |
| LMCA | 52% | 45.2% |
| Epoch Capabilities Index | 156.81 | 153.92 |
| SimpleBench | — | 70% |
| EnigmaEval | 16% | — |
| Thematic Generalization | 80% | — |
| EBR-Bench | 25.4% | — |
| Mystery Game Puzzles | 37% | — |
| Surface Evolver Bench | — | 74.4% |
| ForecastBench | 59.5 | — |
Math GPT-5.4 leads
GPT-5.4: 73.5 (#19), Grok 4.5: 60.9 (#35)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 78.6% | 57.2% |
| FrontierMath Tier 4 | 49% | 24.4% |
| OTIS Mock AIME 2024-2025 | 97.8% | 97.8% |
| ProofBench | 56% | 31% |
| LMArena Math | 1488 | 1459 |
| MathArena Final-Answer Competitions | 83.1% | — |
| FrontierMath (Feb 2025 set) | 47.6% | — |
| FrontierMath Tier 4 (v1) | 27.1% | — |
Knowledge GPT-5.4 leads
GPT-5.4: 65.3 (#14), Grok 4.5: 62.3 (#24)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 93.3% | 93.4% |
| SimpleQA Verified | 45.1% | 48.3% |
| LMArena Expert | 1507 | 1466 |
| Humanity's Last Exam | 36.2% | — |
| Vectara Hallucination Rate | 7% | — |
Multimodal GPT-5.4 leads
GPT-5.4: 43.7 (#20), Grok 4.5: 37.6 (#72)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| LMArena Vision | 1303 | 1288 |
| Blueprint-Bench 2 | 27.1% | 27.3% |
| Furniture Assembly | 37.5% | 22.5% |
| LMArena Document | 1471 | 1452 |
Multilingual GPT-5.4 leads
GPT-5.4: 56.2 (#23), Grok 4.5: 54.4 (#42)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1465 | 1440 |
| LMArena Chinese | 1519 | 1496 |
| LMArena French | 1493 | 1456 |
| LMArena German | 1472 | 1446 |
| LMArena Japanese | 1485 | 1428 |
| LMArena Korean | 1448 | 1404 |
| LMArena Russian | 1480 | 1448 |
| LMArena Spanish | 1454 | 1450 |
Instruction Following GPT-5.4 leads
GPT-5.4: 77.1 (#27), Grok 4.5: 76.0 (#48)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1469 | 1446 |
Long Context GPT-5.4 leads
GPT-5.4: 50.3 (#8), Grok 4.5: 44.8 (#56)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1473 | 1463 |
| CL-bench | 27.9% | — |
| CL-bench Life | 21.7% | — |
Writing & Preference GPT-5.4 leads
GPT-5.4: 71.9 (#17), Grok 4.5: 65.8 (#42)
| Benchmark | GPT-5.4 | Grok 4.5 |
|---|---|---|
| LMArena Text | 1469 | 1448 |
| LMArena Creative Writing | 1439 | 1442 |
| EQ-Bench Creative Writing | 1840 | 1579 |
| LMArena Multi-Turn | 1482 | 1456 |
| EQ-Bench 4 | 1272 | — |
Frequently asked questions
Is GPT-5.4 better than Grok 4.5?
GPT-5.4 is the stronger model overall, scoring 59.4 to 55.0 on the Noometry Index. Grok 4.5 costs 1.9× less per token, which makes it the better buy when GPT-5.4's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 or Grok 4.5?
Grok 4.5 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-5.4 lists at $2.50 and $15.
Is GPT-5.4 or Grok 4.5 better for coding?
They score almost the same on coding (52.6 vs 52.2); test both on your own repository before choosing.
Which has the bigger context window?
GPT-5.4 does, with 1.05M tokens against 500K.
How many benchmarks do GPT-5.4 and Grok 4.5 share?
48 benchmarks have published results for both models. GPT-5.4 has 68 scored results on Noometry and Grok 4.5 has 52.