Model comparison
Grok 4.5 vs Llama 3-70B
Grok 4.5 is the stronger model overall, scoring 55.0 to 28.8 on the Noometry Index.
Last verified . 22 shared benchmarks.
MetaSummary
- They share 22 benchmarks with published results for both. Grok 4.5 scores higher in 9 categories and Llama 3-70B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 4.5 leads 60.9 to 12.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 97.8% for Grok 4.5 and 4.3% for Llama 3-70B.
- Llama 3-70B has downloadable open weights; the other is API-only.
Side by side
| Grok 4.5 | Llama 3-70B | |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 55.0 | 28.8 |
| Released | 2026-07-08 | 2024-04-18 |
| Weights | Proprietary | Open |
| Context window | 500K | — |
| Max output | 500K | — |
| Input $ / M tokens | $2 | — |
| Output $ / M tokens | $6 | — |
| Results tracked | 52 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
Grok 4.5: 52.2 (#35), Llama 3-70B: 35.8 (#218)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| LMArena Coding | 1474 | 1206 |
| DeepSWE | 53.8% | — |
| FrontierCode | 42.4% | — |
| LMArena WebDev | 1553 | — |
| SciCode | 54.1% | — |
| WeirdML | 46.4% | — |
| BigCodeBench Instruct | — | 43.6% |
| BigCodeBench Complete | — | 54.5% |
| ALE-Bench | 1,309 | — |
| HumanEval+ | — | 72% |
| MBPP+ | — | 69% |
Agentic & Tool Use Grok 4.5 leads
Grok 4.5: 44.4 (#17), Llama 3-70B: 21.1 (#139)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| APEX-Agents | 56.2% | — |
| τ²-bench Banking | 47.9% | — |
| Cybench | — | 5% |
| 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: 56.1 (#25), Llama 3-70B: 18.0 (#288)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| Kagi LLM Benchmark | 83.5% | 35.1% |
| LMArena Hard Prompts | 1462 | 1195 |
| DTBench | 96.5% | 54.2% |
| Epoch Capabilities Index | 153.92 | 122.93 |
| ARC-AGI-2 | 52.6% | — |
| SimpleBench | 70% | — |
| NYT Connections (extended) | 79.9% | — |
| ARC-AGI-1 | 87.2% | — |
| CritPt | 15.4% | — |
| Chess Puzzles | 36% | — |
| LMCA | 45.2% | — |
| Surface Evolver Bench | 74.4% | — |
| ForecastBench | — | 57.1 |
| WinoGrande | — | 83.5% |
Math Grok 4.5 leads
Grok 4.5: 60.9 (#35), Llama 3-70B: 12.8 (#305)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 97.8% | 4.3% |
| LMArena Math | 1459 | 1218 |
| FrontierMath (Tiers 1-3) | 57.2% | — |
| FrontierMath Tier 4 | 24.4% | — |
| ProofBench | 31% | — |
| MATH Level 5 | — | 22.6% |
Knowledge Grok 4.5 leads
Grok 4.5: 62.3 (#24), Llama 3-70B: 20.8 (#277)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| GPQA Diamond | 93.4% | 40.6% |
| LMArena Expert | 1466 | 1149 |
| SimpleQA Verified | 48.3% | — |
| MMLU | — | 79.3% |
Multimodal Not comparable
Grok 4.5: 37.6 (#72), Llama 3-70B: —
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| LMArena Vision | 1288 | — |
| Blueprint-Bench 2 | 27.3% | — |
| Furniture Assembly | 22.5% | — |
| LMArena Document | 1452 | — |
Multilingual Grok 4.5 leads
Grok 4.5: 54.4 (#42), Llama 3-70B: 33.6 (#251)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| LMArena Non-English | 1440 | 1142 |
| LMArena Chinese | 1496 | 1114 |
| LMArena French | 1456 | 1232 |
| LMArena German | 1446 | 1169 |
| LMArena Japanese | 1428 | 1017 |
| LMArena Korean | 1404 | 1017 |
| LMArena Russian | 1448 | 1159 |
| LMArena Spanish | 1450 | 1241 |
Instruction Following Grok 4.5 leads
Grok 4.5: 76.0 (#48), Llama 3-70B: 62.5 (#238)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| LMArena Instruction Following | 1446 | 1194 |
Long Context Grok 4.5 leads
Grok 4.5: 44.8 (#56), Llama 3-70B: 35.6 (#240)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| LMArena Longer Query | 1463 | 1174 |
Writing & Preference Grok 4.5 leads
Grok 4.5: 65.8 (#42), Llama 3-70B: 42.8 (#231)
| Benchmark | Grok 4.5 | Llama 3-70B |
|---|---|---|
| LMArena Text | 1448 | 1221 |
| LMArena Creative Writing | 1442 | 1210 |
| LMArena Multi-Turn | 1456 | 1223 |
| EQ-Bench Creative Writing | 1579 | — |
Frequently asked questions
Is Grok 4.5 better than Llama 3-70B?
Grok 4.5 is the stronger model overall, scoring 55.0 to 28.8 on the Noometry Index.
Is Grok 4.5 or Llama 3-70B better for coding?
Grok 4.5 scores higher on coding benchmarks: 52.2 versus 35.8 in the Noometry coding category.
How many benchmarks do Grok 4.5 and Llama 3-70B share?
22 benchmarks have published results for both models. Grok 4.5 has 52 scored results on Noometry and Llama 3-70B has 31.