Model comparison
Grok 3 vs Llama 2-70B
Grok 3 is the stronger model overall, scoring 39.9 to 24.4 on the Noometry Index.
Last verified . 21 shared benchmarks.
MetaSummary
- They share 21 benchmarks with published results for both. Grok 3 scores higher in 7 categories and Llama 2-70B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok 3 leads 46.2 to 7.4.
- The biggest single-benchmark swing is MATH Level 5: 88.7% for Grok 3 and 3.3% for Llama 2-70B.
- Llama 2-70B has downloadable open weights; the other is API-only.
Side by side
| Grok 3 | Llama 2-70B | |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 39.9 | 24.4 |
| Released | 2025-04-09 | 2023-07-18 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 40 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 3 leads
Grok 3: 41.9 (#115), Llama 2-70B: 31.4 (#286)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1432 | 1079 |
| Aider Polyglot | 53.3% | — |
| WeirdML | 37.2% | — |
Agentic & Tool Use Not comparable
Grok 3: 30.5 (#76), Llama 2-70B: —
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| BALROG | 29.5% | — |
Reasoning Too close to call
Grok 3: 13.7 (#333), Llama 2-70B: 14.4 (#325)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1434 | 1073 |
| Epoch Capabilities Index | 138.33 | 113.79 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 36.1% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 5.5% | — |
| DTBench | — | 41.6% |
| BIG-Bench Hard | — | 64.9% |
| CommonsenseQA 2.0 | — | 50% |
| ForecastBench | — | 51.4 |
| HellaSwag | — | 85.3% |
| LAMBADA | — | 78.9% |
| PIQA | — | 82.8% |
| WinoGrande | — | 80.2% |
Math Grok 3 leads
Grok 3: 38.0 (#145), Llama 2-70B: 8.1 (#326)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 55.6% | 0% |
| LMArena Math | 1391 | 1091 |
| MATH Level 5 | 88.7% | 3.3% |
| Omni-MATH | 46.4% | — |
| FrontierMath (Feb 2025 set) | 3.8% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
| GSM8K | — | 69.6% |
Knowledge Grok 3 leads
Grok 3: 46.2 (#82), Llama 2-70B: 7.4 (#310)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| GPQA Diamond | 75.8% | 26.3% |
| LMArena Expert | 1421 | 1039 |
| MMLU-Pro | 78.8% | — |
| Confabulations | 14.2% | — |
| Vectara Hallucination Rate | 5.8% | — |
| GPQA (HELM) | 65% | — |
| ARC (AI2) Challenge | — | 78.3% |
| BoolQ | — | 88.6% |
| MMLU | — | 69.9% |
| OpenBookQA | — | 60.2% |
| TriviaQA | — | 87.6% |
Multilingual Grok 3 leads
Grok 3: 52.3 (#87), Llama 2-70B: 27.7 (#274)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 1410 | 1045 |
| LMArena Chinese | 1448 | 995 |
| LMArena French | 1460 | 1090 |
| LMArena German | 1431 | 1041 |
| LMArena Japanese | 1387 | 927 |
| LMArena Korean | 1373 | 964 |
| LMArena Russian | 1416 | 1083 |
| LMArena Spanish | 1417 | 1143 |
Instruction Following Grok 3 leads
Grok 3: 75.0 (#73), Llama 2-70B: 54.9 (#278)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1409 | 1071 |
| IFEval | 88.4% | — |
Long Context Grok 3 leads
Grok 3: 38.7 (#192), Llama 2-70B: 32.3 (#270)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1439 | 1062 |
| Fiction.LiveBench | 58.3% | — |
Writing & Preference Grok 3 leads
Grok 3: 55.8 (#141), Llama 2-70B: 32.3 (#279)
| Benchmark | Grok 3 | Llama 2-70B |
|---|---|---|
| LMArena Text | 1426 | 1115 |
| LMArena Creative Writing | 1414 | 1075 |
| LMArena Multi-Turn | 1425 | 1088 |
| Short-Story Creative Writing | 76.4% | — |
| EQ-Bench Creative Writing | 1186 | — |
| WildBench | 84.9% | — |
Frequently asked questions
Is Grok 3 better than Llama 2-70B?
Grok 3 is the stronger model overall, scoring 39.9 to 24.4 on the Noometry Index.
Is Grok 3 or Llama 2-70B better for coding?
Grok 3 scores higher on coding benchmarks: 41.9 versus 31.4 in the Noometry coding category.
How many benchmarks do Grok 3 and Llama 2-70B share?
21 benchmarks have published results for both models. Grok 3 has 40 scored results on Noometry and Llama 2-70B has 35.