Model comparison
Grok 3 vs Llama 3.2 1B
Grok 3 is the stronger model overall, scoring 39.6 to 19.9 on the Noometry Index.
Last verified . 18 shared benchmarks.
MetaSummary
- They share 18 benchmarks with published results for both. Grok 3 scores higher in 8 categories and Llama 3.2 1B in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok 3 leads 46.0 to 7.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 55.6% for Grok 3 and 0.6% for Llama 3.2 1B.
- Llama 3.2 1B has downloadable open weights; the other is API-only.
Side by side
| Grok 3 | Llama 3.2 1B | |
|---|---|---|
| Provider | xAI | Meta |
| Noometry Index | 39.6 | 19.9 |
| Released | 2025-04-09 | 2024-09-24 |
| Weights | Proprietary | Open |
| Context window | — | 60K |
| Max output | — | 54K |
| Input $ / M tokens | — | $0.027 |
| Output $ / M tokens | — | $0.20 |
| Results tracked | 40 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 3 leads
Grok 3: 41.5 (#113), Llama 3.2 1B: 20.6 (#333)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| LMArena Coding | 1432 | 1071 |
| Aider Polyglot | 53.3% | — |
| WeirdML | 37.2% | — |
| BigCodeBench Instruct | — | 8.2% |
| BigCodeBench Complete | — | 11.3% |
Agentic & Tool Use Grok 3 leads
Grok 3: 30.2 (#69), Llama 3.2 1B: 14.5 (#142)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| BALROG | 29.5% | 6.6% |
| Berkeley Function Calling Leaderboard | — | 10.8% |
Reasoning Llama 3.2 1B leads
Grok 3: 13.8 (#322), Llama 3.2 1B: 15.6 (#304)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| LMArena Hard Prompts | 1433 | 1044 |
| Epoch Capabilities Index | 138.34 | 101.95 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 36.1% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 5.5% | — |
| Chess Puzzles | — | 0% |
Math Grok 3 leads
Grok 3: 36.2 (#122), Llama 3.2 1B: 9.7 (#308)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 55.6% | 0.6% |
| LMArena Math | 1390 | 1086 |
| Omni-MATH | 46.4% | — |
| MATH Level 5 | 88.7% | — |
| FrontierMath (Feb 2025 set) | 3.8% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Grok 3 leads
Grok 3: 46.0 (#76), Llama 3.2 1B: 7.2 (#307)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| GPQA Diamond | 75.8% | 23.9% |
| LMArena Expert | 1420 | 1007 |
| MMLU-Pro | 78.8% | — |
| Confabulations | 14.2% | — |
| Vectara Hallucination Rate | 5.8% | — |
| GPQA (HELM) | 65% | — |
Multilingual Grok 3 leads
Grok 3: 52.5 (#81), Llama 3.2 1B: 23.9 (#287)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| LMArena Non-English | 1411 | 973 |
| LMArena Chinese | 1448 | 958 |
| LMArena German | 1431 | 1014 |
| LMArena Russian | 1416 | 941 |
| LMArena French | 1460 | — |
| LMArena Japanese | 1389 | — |
| LMArena Korean | 1374 | — |
| LMArena Spanish | 1417 | — |
Instruction Following Grok 3 leads
Grok 3: 75.2 (#69), Llama 3.2 1B: 52.6 (#285)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| LMArena Instruction Following | 1409 | 1031 |
| IFEval | 88.4% | — |
Long Context Grok 3 leads
Grok 3: 38.7 (#187), Llama 3.2 1B: 31.9 (#269)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| LMArena Longer Query | 1439 | 1050 |
| Fiction.LiveBench | 58.3% | — |
Writing & Preference Grok 3 leads
Grok 3: 56.0 (#136), Llama 3.2 1B: 21.4 (#305)
| Benchmark | Grok 3 | Llama 3.2 1B |
|---|---|---|
| LMArena Text | 1426 | 1055 |
| LMArena Creative Writing | 1414 | 1033 |
| EQ-Bench Creative Writing | 1186 | 200 |
| LMArena Multi-Turn | 1424 | 1030 |
| Short-Story Creative Writing | 76.4% | — |
| WildBench | 84.9% | — |
Frequently asked questions
Is Grok 3 better than Llama 3.2 1B?
Grok 3 is the stronger model overall, scoring 39.6 to 19.9 on the Noometry Index.
Is Grok 3 or Llama 3.2 1B better for coding?
Grok 3 scores higher on coding benchmarks: 41.5 versus 20.6 in the Noometry coding category.
How many benchmarks do Grok 3 and Llama 3.2 1B share?
18 benchmarks have published results for both models. Grok 3 has 40 scored results on Noometry and Llama 3.2 1B has 22.