Model comparison
GPT-4o vs Llama 2-70B
GPT-4o is the stronger model overall, scoring 28.6 to 24.4 on the Noometry Index.
Last verified . 24 shared benchmarks.
MetaSummary
- They share 24 benchmarks with published results for both. GPT-4o scores higher in 6 categories and Llama 2-70B in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-4o leads 28.8 to 7.4.
- The biggest single-benchmark swing is MATH Level 5: 53.3% for GPT-4o and 3.3% for Llama 2-70B.
- Llama 2-70B has downloadable open weights; the other is API-only.
Side by side
| GPT-4o | Llama 2-70B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 28.6 | 24.4 |
| Released | 2024-05-13 | 2023-07-18 |
| Weights | Proprietary | Open |
| Context window | 128K | — |
| Max output | 16K | — |
| Input $ / M tokens | $2.50 | — |
| Output $ / M tokens | $10 | — |
| Results tracked | 72 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-70B leads
GPT-4o: 24.8 (#328), Llama 2-70B: 31.4 (#286)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1297 | 1079 |
| SWE-bench Verified | 31% | — |
| SWE-bench Verified (bash only) | 21.6% | — |
| Aider Polyglot | 45.3% | — |
| GSO | 0% | — |
| WeirdML | 25.1% | — |
| BigCodeBench Instruct | 51.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 61.1% | — |
| CadEval | 26% | — |
| HumanEval+ | 87.2% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use Not comparable
GPT-4o: 21.0 (#141), Llama 2-70B: —
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| GDPval | 9.9% | — |
| TheAgentCompany | 8.6% | — |
| Cybench | 12.5% | — |
| BALROG | 32.3% | — |
| LMArena Search | 1006 | — |
| METR Time Horizons | 40.8% | — |
Reasoning Llama 2-70B leads
GPT-4o: 9.4 (#343), Llama 2-70B: 14.4 (#325)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1281 | 1073 |
| DTBench | 64.5% | 41.6% |
| Epoch Capabilities Index | 128.97 | 113.79 |
| ForecastBench | 57.7 | 51.4 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 17.8% | — |
| ARC-AGI-1 | 4.5% | — |
| CritPt | 0% | — |
| Chess Puzzles | 13% | — |
| EnigmaEval | 0.8% | — |
| LiveBench Reasoning | 55.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 16.6% | — |
| BIG-Bench Hard | — | 64.9% |
| CommonsenseQA 2.0 | — | 50% |
| HellaSwag | — | 85.3% |
| LAMBADA | — | 78.9% |
| LiveBench | 55.3% | — |
| PIQA | — | 82.8% |
| WinoGrande | — | 80.2% |
Math GPT-4o leads
GPT-4o: 10.6 (#312), Llama 2-70B: 8.1 (#326)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 6.4% | 0% |
| LMArena Math | 1285 | 1091 |
| MATH Level 5 | 53.3% | 3.3% |
| FrontierMath (Tiers 1-3) | 0.4% | — |
| Omni-MATH | 29.3% | — |
| LiveBench Math | 49.5% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
| GSM8K | — | 69.6% |
Knowledge GPT-4o leads
GPT-4o: 28.8 (#242), Llama 2-70B: 7.4 (#310)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| GPQA Diamond | 49.2% | 26.3% |
| LMArena Expert | 1250 | 1039 |
| MMLU | 88.1% | 69.9% |
| Humanity's Last Exam | 2.7% | — |
| SimpleQA Verified | 26% | — |
| MMLU-Pro | 71.3% | — |
| Confabulations | 15.3% | — |
| Vectara Hallucination Rate | 9.6% | — |
| GPQA (HELM) | 52% | — |
| ARC (AI2) Challenge | — | 78.3% |
| BoolQ | — | 88.6% |
| OpenBookQA | — | 60.2% |
| TriviaQA | — | 87.6% |
Multimodal Not comparable
GPT-4o: 34.5 (#91), Llama 2-70B: —
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| LMArena Vision | 1137 | — |
| Video-MME | 71.9% | — |
| GeoBench | 71% | — |
| VPCT | 40% | — |
| ScienceQA | 88.5% | — |
Multilingual GPT-4o leads
GPT-4o: 43.2 (#186), Llama 2-70B: 27.7 (#274)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 1283 | 1045 |
| LMArena Chinese | 1277 | 995 |
| LMArena French | 1304 | 1090 |
| LMArena German | 1282 | 1041 |
| LMArena Japanese | 1257 | 927 |
| LMArena Korean | 1234 | 964 |
| LMArena Russian | 1286 | 1083 |
| LMArena Spanish | 1292 | 1143 |
Instruction Following GPT-4o leads
GPT-4o: 66.6 (#207), Llama 2-70B: 54.9 (#278)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1278 | 1071 |
| LiveBench Instruction Following | 68.6% | — |
| IFEval | 81.7% | — |
Long Context GPT-4o leads
GPT-4o: 39.4 (#179), Llama 2-70B: 32.3 (#270)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1289 | 1062 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference GPT-4o leads
GPT-4o: 52.6 (#166), Llama 2-70B: 32.3 (#279)
| Benchmark | GPT-4o | Llama 2-70B |
|---|---|---|
| LMArena Text | 1300 | 1115 |
| LMArena Creative Writing | 1292 | 1075 |
| LMArena Multi-Turn | 1302 | 1088 |
| Short-Story Creative Writing | 81.8% | — |
| WildBench | 82.8% | — |
| LiveBench Language | 47.6% | — |
Frequently asked questions
Is GPT-4o better than Llama 2-70B?
GPT-4o is the stronger model overall, scoring 28.6 to 24.4 on the Noometry Index.
Is GPT-4o or Llama 2-70B better for coding?
Llama 2-70B scores higher on coding benchmarks: 31.4 versus 24.8 in the Noometry coding category.
How many benchmarks do GPT-4o and Llama 2-70B share?
24 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and Llama 2-70B has 35.