Model comparison
GPT-5.4 nano vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 41.9 on the Noometry Index. GPT-5.4 nano costs 6.5× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Last verified . 35 shared benchmarks.
Summary
- They share 35 benchmarks with published results for both. GPT-5.4 nano scores higher in 0 categories and Grok 4.5 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.5 leads 56.0 to 23.8.
- The biggest single-benchmark swing is ARC-AGI-2: 5.7% for GPT-5.4 nano and 52.6% for Grok 4.5.
- GPT-5.4 nano is cheaper at $0.20 / $1.25 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 nano | Grok 4.5 | |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 41.9 | 55.0 |
| Released | 2026-03-17 | 2026-07-08 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $0.20 | $2 |
| Output $ / M tokens | $1.25 | $6 |
| Results tracked | 40 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
GPT-5.4 nano: 43.1 (#82), Grok 4.5: 51.7 (#34)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| SciCode | 46.9% | 54.1% |
| WeirdML | 49.2% | 46.4% |
| LMArena Coding | 1404 | 1473 |
| ALE-Bench | 1,005 | 1,309 |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.4% |
| LMArena WebDev | — | 1552 |
Agentic & Tool Use Not comparable
GPT-5.4 nano: —, Grok 4.5: 44.4 (#15)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| APEX-Agents | — | 56.2% |
| τ²-bench Banking | — | 47.9% |
| 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 nano: 23.8 (#159), Grok 4.5: 56.0 (#23)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| ARC-AGI-2 | 5.7% | 52.6% |
| Kagi LLM Benchmark | 39.7% | 83.5% |
| ARC-AGI-1 | 51.5% | 87.2% |
| CritPt | 9.3% | 15.4% |
| Chess Puzzles | 30% | 36% |
| LMArena Hard Prompts | 1380 | 1462 |
| DTBench | 80.3% | 96.5% |
| LMCA | 36.9% | 45.2% |
| Epoch Capabilities Index | 145.78 | 154.02 |
| SimpleBench | — | 70% |
| NYT Connections (extended) | — | 79.9% |
| Mystery Game Puzzles | 9% | — |
| Surface Evolver Bench | — | 74.4% |
| ForecastBench | 57.3 | — |
Math Grok 4.5 leads
GPT-5.4 nano: 40.7 (#76), Grok 4.5: 60.7 (#29)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 44.9% | 57.2% |
| FrontierMath Tier 4 | 12.2% | 24.4% |
| OTIS Mock AIME 2024-2025 | 87.8% | 97.8% |
| ProofBench | 5% | 31% |
| LMArena Math | 1406 | 1457 |
| FrontierMath (Feb 2025 set) | 25.9% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge Grok 4.5 leads
GPT-5.4 nano: 41.8 (#94), Grok 4.5: 62.3 (#21)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 78.5% | 93.4% |
| SimpleQA Verified | 11.7% | 48.3% |
| LMArena Expert | 1397 | 1466 |
| Vectara Hallucination Rate | 3.1% | — |
Multimodal Grok 4.5 leads
GPT-5.4 nano: 36.0 (#73), Grok 4.5: 37.8 (#63)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| LMArena Vision | 1196 | 1287 |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.5 leads
GPT-5.4 nano: 48.8 (#136), Grok 4.5: 54.5 (#43)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1359 | 1439 |
| LMArena Chinese | 1391 | 1497 |
| LMArena French | 1396 | 1458 |
| LMArena German | 1368 | 1448 |
| LMArena Japanese | 1342 | 1425 |
| LMArena Korean | 1321 | 1402 |
| LMArena Russian | 1363 | 1447 |
| LMArena Spanish | 1368 | 1452 |
Instruction Following Grok 4.5 leads
GPT-5.4 nano: 72.1 (#139), Grok 4.5: 76.3 (#45)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1362 | 1446 |
Long Context Grok 4.5 leads
GPT-5.4 nano: 41.6 (#131), Grok 4.5: 44.8 (#54)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1365 | 1463 |
Writing & Preference Grok 4.5 leads
GPT-5.4 nano: 55.9 (#137), Grok 4.5: 65.9 (#40)
| Benchmark | GPT-5.4 nano | Grok 4.5 |
|---|---|---|
| LMArena Text | 1372 | 1448 |
| LMArena Creative Writing | 1313 | 1442 |
| LMArena Multi-Turn | 1382 | 1457 |
| EQ-Bench Creative Writing | — | 1579 |
Frequently asked questions
Is GPT-5.4 nano better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 41.9 on the Noometry Index. GPT-5.4 nano costs 6.5× 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 nano or Grok 4.5?
GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; Grok 4.5 lists at $2 and $6.
Is GPT-5.4 nano or Grok 4.5 better for coding?
Grok 4.5 scores higher on coding benchmarks: 51.7 versus 43.1 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 nano and Grok 4.5 share?
35 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and Grok 4.5 has 52.