Model comparison
GPT-4.1 vs Llama 3-70B
GPT-4.1 is the stronger model overall, scoring 35.9 to 28.8 on the Noometry Index.
Last verified . 24 shared benchmarks.
MetaSummary
- They share 24 benchmarks with published results for both. GPT-4.1 scores higher in 7 categories and Llama 3-70B in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-4.1 leads 37.1 to 20.8.
- The biggest single-benchmark swing is MATH Level 5: 83% for GPT-4.1 and 22.6% for Llama 3-70B.
- Llama 3-70B has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Llama 3-70B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 35.9 | 28.8 |
| Released | 2025-04-14 | 2024-04-18 |
| Weights | Proprietary | Open |
| Context window | 1.05M | — |
| Max output | 33K | — |
| Input $ / M tokens | $2 | — |
| Output $ / M tokens | $8 | — |
| Results tracked | 52 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3-70B leads
GPT-4.1: 34.4 (#238), Llama 3-70B: 35.8 (#218)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| LMArena Coding | 1391 | 1206 |
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| WeirdML | 39% | — |
| BigCodeBench Instruct | — | 43.6% |
| BigCodeBench Complete | — | 54.5% |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
| HumanEval+ | — | 72% |
| MBPP+ | — | 69% |
Agentic & Tool Use GPT-4.1 leads
GPT-4.1: 34.7 (#43), Llama 3-70B: 21.1 (#139)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | — |
| Cybench | — | 5% |
Reasoning Llama 3-70B leads
GPT-4.1: 11.7 (#339), Llama 3-70B: 18.0 (#288)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| Kagi LLM Benchmark | 52.3% | 35.1% |
| LMArena Hard Prompts | 1384 | 1195 |
| DTBench | 68.3% | 54.2% |
| Epoch Capabilities Index | 136.78 | 122.93 |
| ForecastBench | 61.5 | 57.1 |
| ARC-AGI-2 | 0.4% | — |
| SimpleBench | 27% | — |
| ARC-AGI-1 | 5.5% | — |
| Chess Puzzles | 6% | — |
| EnigmaEval | 2.2% | — |
| LMCA | 25.6% | — |
| WinoGrande | — | 83.5% |
Math GPT-4.1 leads
GPT-4.1: 22.3 (#280), Llama 3-70B: 12.8 (#305)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 38.3% | 4.3% |
| LMArena Math | 1370 | 1218 |
| MATH Level 5 | 83% | 22.6% |
| FrontierMath (Tiers 1-3) | 6% | — |
| Omni-MATH | 47.1% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge GPT-4.1 leads
GPT-4.1: 37.1 (#160), Llama 3-70B: 20.8 (#277)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| GPQA Diamond | 66.9% | 40.6% |
| LMArena Expert | 1364 | 1149 |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| MMLU-Pro | 81.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
| MMLU | — | 79.3% |
Multimodal Not comparable
GPT-4.1: 38.2 (#67), Llama 3-70B: —
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 72% | — |
Multilingual GPT-4.1 leads
GPT-4.1: 49.4 (#133), Llama 3-70B: 33.6 (#251)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| LMArena Non-English | 1370 | 1142 |
| LMArena Chinese | 1382 | 1114 |
| LMArena French | 1382 | 1232 |
| LMArena German | 1381 | 1169 |
| LMArena Japanese | 1319 | 1017 |
| LMArena Korean | 1339 | 1017 |
| LMArena Russian | 1377 | 1159 |
| LMArena Spanish | 1376 | 1241 |
Instruction Following GPT-4.1 leads
GPT-4.1: 71.3 (#153), Llama 3-70B: 62.5 (#238)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| LMArena Instruction Following | 1367 | 1194 |
| IFEval | 83.8% | — |
Long Context GPT-4.1 leads
GPT-4.1: 40.0 (#163), Llama 3-70B: 35.6 (#240)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| LMArena Longer Query | 1385 | 1174 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-4.1 leads
GPT-4.1: 57.6 (#125), Llama 3-70B: 42.8 (#231)
| Benchmark | GPT-4.1 | Llama 3-70B |
|---|---|---|
| LMArena Text | 1383 | 1221 |
| LMArena Creative Writing | 1363 | 1210 |
| LMArena Multi-Turn | 1398 | 1223 |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is GPT-4.1 better than Llama 3-70B?
GPT-4.1 is the stronger model overall, scoring 35.9 to 28.8 on the Noometry Index.
Is GPT-4.1 or Llama 3-70B better for coding?
Llama 3-70B scores higher on coding benchmarks: 35.8 versus 34.4 in the Noometry coding category.
How many benchmarks do GPT-4.1 and Llama 3-70B share?
24 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Llama 3-70B has 31.