Model comparison
GPT-5.4 mini vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 44.8 on the Noometry Index. GPT-5.4 mini costs 1.8× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Last verified . 39 shared benchmarks.
Summary
- They share 39 benchmarks with published results for both. GPT-5.4 mini scores higher in 1 category and Grok 4.5 in 9 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.5 leads 56.0 to 30.3.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 37.9% for GPT-5.4 mini and 83.5% for Grok 4.5.
- GPT-5.4 mini is cheaper at $0.75 / $4.50 per million input/output tokens, against $2 / $6 for Grok 4.5.
- Grok 4.5 accepts more context: 500K tokens versus 400K.
Side by side
| GPT-5.4 mini | Grok 4.5 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 44.8 | 55.0 |
| Released | 2026-03-17 | 2026-07-08 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $0.75 | $2 |
| Output $ / M tokens | $4.50 | $6 |
| Results tracked | 46 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
GPT-5.4 mini: 45.0 (#67), Grok 4.5: 51.7 (#34)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| FrontierCode | 27% | 42.4% |
| LMArena WebDev | 1397 | 1552 |
| SciCode | 49.9% | 54.1% |
| WeirdML | 60.3% | 46.4% |
| LMArena Coding | 1438 | 1473 |
| ALE-Bench | 1,189 | 1,309 |
| DeepSWE | — | 53.8% |
Agentic & Tool Use Grok 4.5 leads
GPT-5.4 mini: 29.6 (#74), Grok 4.5: 44.4 (#15)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| APEX-Agents | — | 56.2% |
| τ²-bench Banking | — | 47.9% |
| DeepResearch Bench | 36.3% | — |
| PostTrainBench | — | 23.4% |
| GBAEval | — | 65.4% |
| GDP.pdf | — | 14% |
| LMArena Search | — | 1213 |
| Vending-Bench 2 | — | 3,887 |
Reasoning Grok 4.5 leads
GPT-5.4 mini: 30.3 (#77), Grok 4.5: 56.0 (#23)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| ARC-AGI-2 | 18.9% | 52.6% |
| Kagi LLM Benchmark | 37.9% | 83.5% |
| NYT Connections (extended) | 61.8% | 79.9% |
| ARC-AGI-1 | 63.7% | 87.2% |
| CritPt | 10% | 15.4% |
| Chess Puzzles | 24% | 36% |
| LMArena Hard Prompts | 1424 | 1462 |
| DTBench | 80% | 96.5% |
| LMCA | 40.8% | 45.2% |
| Epoch Capabilities Index | 148.83 | 154.02 |
| SimpleBench | — | 70% |
| Thematic Generalization | 61.7% | — |
| Mystery Game Puzzles | 11% | — |
| Surface Evolver Bench | — | 74.4% |
| ForecastBench | 57 | — |
Math Grok 4.5 leads
GPT-5.4 mini: 45.4 (#63), Grok 4.5: 60.7 (#29)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 57.2% |
| FrontierMath Tier 4 | 9.8% | 24.4% |
| OTIS Mock AIME 2024-2025 | 88.9% | 97.8% |
| ProofBench | 21% | 31% |
| LMArena Math | 1418 | 1457 |
| FrontierMath (Feb 2025 set) | 28.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Grok 4.5 leads
GPT-5.4 mini: 51.4 (#62), Grok 4.5: 62.3 (#21)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 86.9% | 93.4% |
| SimpleQA Verified | 29.4% | 48.3% |
| LMArena Expert | 1434 | 1466 |
| Vectara Hallucination Rate | 5.5% | — |
Multimodal GPT-5.4 mini leads
GPT-5.4 mini: 39.0 (#51), Grok 4.5: 37.8 (#63)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| LMArena Vision | 1245 | 1287 |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.5 leads
GPT-5.4 mini: 52.1 (#94), Grok 4.5: 54.5 (#43)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1405 | 1439 |
| LMArena Chinese | 1446 | 1497 |
| LMArena French | 1440 | 1458 |
| LMArena German | 1407 | 1448 |
| LMArena Japanese | 1375 | 1425 |
| LMArena Korean | 1369 | 1402 |
| LMArena Russian | 1417 | 1447 |
| LMArena Spanish | 1403 | 1452 |
Instruction Following Grok 4.5 leads
GPT-5.4 mini: 74.3 (#99), Grok 4.5: 76.3 (#45)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1446 |
Long Context Grok 4.5 leads
GPT-5.4 mini: 42.9 (#107), Grok 4.5: 44.8 (#54)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1407 | 1463 |
Writing & Preference Grok 4.5 leads
GPT-5.4 mini: 64.2 (#56), Grok 4.5: 65.9 (#40)
| Benchmark | GPT-5.4 mini | Grok 4.5 |
|---|---|---|
| LMArena Text | 1411 | 1448 |
| LMArena Creative Writing | 1370 | 1442 |
| EQ-Bench Creative Writing | 1665 | 1579 |
| LMArena Multi-Turn | 1429 | 1457 |
Frequently asked questions
Is GPT-5.4 mini better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 44.8 on the Noometry Index. GPT-5.4 mini costs 1.8× 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-5.4 mini or Grok 4.5?
GPT-5.4 mini is cheaper. It lists at $0.75 per million input tokens and $4.50 per million output tokens; Grok 4.5 lists at $2 and $6.
Is GPT-5.4 mini or Grok 4.5 better for coding?
Grok 4.5 scores higher on coding benchmarks: 51.7 versus 45.0 in the Noometry coding category.
Which has the bigger context window?
Grok 4.5 does, with 500K tokens against 400K.
How many benchmarks do GPT-5.4 mini and Grok 4.5 share?
39 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and Grok 4.5 has 52.