Model comparison
Grok 4.5 vs Llama 4 Maverick
Grok 4.5 is the stronger model overall, scoring 55.0 to 30.6 on the Noometry Index. Llama 4 Maverick costs 9.9× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Last verified . 33 shared benchmarks.
MetaSummary
- They share 33 benchmarks with published results for both. Grok 4.5 scores higher in 10 categories and Llama 4 Maverick in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.5 leads 56.0 to 10.1.
- The biggest single-benchmark swing is ARC-AGI-1: 87.2% for Grok 4.5 and 4.4% for Llama 4 Maverick.
- Llama 4 Maverick is cheaper at $0.19 / $0.65 per million input/output tokens, against $2 / $6 for Grok 4.5.
- Grok 4.5 accepts more context: 500K tokens versus 128K.
- Llama 4 Maverick has downloadable open weights; the other is API-only.
Side by side
| Grok 4.5 | Llama 4 Maverick | |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 55.0 | 30.6 |
| Released | 2026-07-08 | 2025-04-05 |
| Weights | Proprietary | Open |
| Context window | 500K | 128K |
| Max output | 500K | 4K |
| Input $ / M tokens | $2 | $0.19 |
| Output $ / M tokens | $6 | $0.65 |
| Results tracked | 52 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
Grok 4.5: 51.7 (#34), Llama 4 Maverick: 26.3 (#318)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| SciCode | 54.1% | 33.1% |
| WeirdML | 46.4% | 24.5% |
| LMArena Coding | 1473 | 1302 |
| ALE-Bench | 1,309 | 172.97 |
| DeepSWE | 53.8% | — |
| FrontierCode | 42.4% | — |
| SWE-bench Verified (bash only) | — | 21% |
| Aider Polyglot | — | 15.6% |
| LMArena WebDev | 1552 | — |
| BigCodeBench Instruct | — | 49.7% |
| BigCodeBench Complete | — | 61.4% |
Agentic & Tool Use Grok 4.5 leads
Grok 4.5: 44.4 (#15), Llama 4 Maverick: 27.9 (#86)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| APEX-Agents | 56.2% | — |
| Berkeley Function Calling Leaderboard | — | 37.3% |
| 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
Grok 4.5: 56.0 (#23), Llama 4 Maverick: 10.1 (#337)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| ARC-AGI-2 | 52.6% | 0% |
| SimpleBench | 70% | 27.7% |
| Kagi LLM Benchmark | 83.5% | 55.9% |
| NYT Connections (extended) | 79.9% | 8% |
| ARC-AGI-1 | 87.2% | 4.4% |
| CritPt | 15.4% | 0% |
| LMArena Hard Prompts | 1462 | 1281 |
| DTBench | 96.5% | 61.9% |
| LMCA | 45.2% | 15.9% |
| Epoch Capabilities Index | 154.02 | 132.2 |
| Chess Puzzles | 36% | — |
| EnigmaEval | — | 0.6% |
| Surface Evolver Bench | 74.4% | — |
| ForecastBench | — | 57.5 |
Math Grok 4.5 leads
Grok 4.5: 60.7 (#29), Llama 4 Maverick: 24.6 (#256)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 97.8% | 20.6% |
| LMArena Math | 1457 | 1299 |
| FrontierMath (Tiers 1-3) | 57.2% | — |
| FrontierMath Tier 4 | 24.4% | — |
| ProofBench | 31% | — |
| Omni-MATH | — | 42.2% |
| MATH Level 5 | — | 73% |
| FrontierMath (Feb 2025 set) | — | 0.7% |
Knowledge Grok 4.5 leads
Grok 4.5: 62.3 (#21), Llama 4 Maverick: 33.3 (#196)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 93.4% | 67% |
| LMArena Expert | 1466 | 1259 |
| Humanity's Last Exam | — | 5.7% |
| SimpleQA Verified | 48.3% | — |
| MMLU-Pro | — | 81% |
| Confabulations | — | 22.6% |
| Vectara Hallucination Rate | — | 8.2% |
| GPQA (HELM) | — | 65% |
Multimodal Grok 4.5 leads
Grok 4.5: 37.8 (#63), Llama 4 Maverick: 30.9 (#100)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | 1287 | 1141 |
| GeoBench | — | 52% |
| Blueprint-Bench 2 | 27.3% | — |
| Furniture Assembly | 22.5% | — |
| LMArena Document | 1452 | — |
| SpatialViz-Bench | — | 31.8% |
Multilingual Grok 4.5 leads
Grok 4.5: 54.5 (#43), Llama 4 Maverick: 42.4 (#190)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1439 | 1269 |
| LMArena Chinese | 1497 | 1277 |
| LMArena French | 1458 | 1258 |
| LMArena German | 1448 | 1291 |
| LMArena Japanese | 1425 | 1207 |
| LMArena Korean | 1402 | 1203 |
| LMArena Russian | 1447 | 1286 |
| LMArena Spanish | 1452 | 1293 |
Instruction Following Grok 4.5 leads
Grok 4.5: 76.3 (#45), Llama 4 Maverick: 71.9 (#140)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Instruction Following | 1446 | 1267 |
| IFEval | — | 90.8% |
Long Context Grok 4.5 leads
Grok 4.5: 44.8 (#54), Llama 4 Maverick: 31.4 (#274)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Longer Query | 1463 | 1280 |
| Fiction.LiveBench | — | 46.2% |
Writing & Preference Grok 4.5 leads
Grok 4.5: 65.9 (#40), Llama 4 Maverick: 39.0 (#247)
| Benchmark | Grok 4.5 | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1448 | 1287 |
| LMArena Creative Writing | 1442 | 1266 |
| EQ-Bench Creative Writing | 1579 | 860 |
| LMArena Multi-Turn | 1457 | 1289 |
| Short-Story Creative Writing | — | 62% |
| WildBench | — | 80% |
Frequently asked questions
Is Grok 4.5 better than Llama 4 Maverick?
Grok 4.5 is the stronger model overall, scoring 55.0 to 30.6 on the Noometry Index. Llama 4 Maverick costs 9.9× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Which is cheaper, Grok 4.5 or Llama 4 Maverick?
Llama 4 Maverick is cheaper. It lists at $0.19 per million input tokens and $0.65 per million output tokens; Grok 4.5 lists at $2 and $6.
Is Grok 4.5 or Llama 4 Maverick better for coding?
Grok 4.5 scores higher on coding benchmarks: 51.7 versus 26.3 in the Noometry coding category.
Which has the bigger context window?
Grok 4.5 does, with 500K tokens against 128K.
How many benchmarks do Grok 4.5 and Llama 4 Maverick share?
33 benchmarks have published results for both models. Grok 4.5 has 52 scored results on Noometry and Llama 4 Maverick has 54.