Model comparison
GPT-4.5 vs Llama 2-70B
GPT-4.5 is the stronger model overall, scoring 37.2 to 24.4 on the Noometry Index.
Last verified . 21 shared benchmarks.
MetaSummary
- They share 21 benchmarks with published results for both. GPT-4.5 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 GPT-4.5 leads 32.5 to 7.4.
- The biggest single-benchmark swing is MATH Level 5: 78.6% for GPT-4.5 and 3.3% for Llama 2-70B.
- Llama 2-70B has downloadable open weights; the other is API-only.
Side by side
| GPT-4.5 | Llama 2-70B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 37.2 | 24.4 |
| Released | 2025-02-27 | 2023-07-18 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 42 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4.5 leads
GPT-4.5: 42.2 (#109), Llama 2-70B: 31.4 (#286)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1396 | 1079 |
| Aider Polyglot | 44.9% | — |
| WeirdML | 39.4% | — |
| LiveBench Coding | 75.2% | — |
Agentic & Tool Use Not comparable
GPT-4.5: 27.9 (#97), Llama 2-70B: —
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| Cybench | 17.5% | — |
Reasoning Too close to call
GPT-4.5: 13.9 (#330), Llama 2-70B: 14.4 (#325)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1403 | 1073 |
| Epoch Capabilities Index | 136.74 | 113.79 |
| ForecastBench | 61.7 | 51.4 |
| ARC-AGI-2 | 0.8% | — |
| SimpleBench | 34.5% | — |
| ARC-AGI-1 | 10.3% | — |
| EnigmaEval | 3.2% | — |
| LiveBench Reasoning | 71.1% | — |
| DTBench | — | 41.6% |
| LiveBench Data Analysis | 64.3% | — |
| BIG-Bench Hard | — | 64.9% |
| CommonsenseQA 2.0 | — | 50% |
| HellaSwag | — | 85.3% |
| LAMBADA | — | 78.9% |
| LiveBench | 69% | — |
| PIQA | — | 82.8% |
| WinoGrande | — | 80.2% |
Math GPT-4.5 leads
GPT-4.5: 32.6 (#211), Llama 2-70B: 8.1 (#326)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 0% |
| LMArena Math | 1412 | 1091 |
| MATH Level 5 | 78.6% | 3.3% |
| LiveBench Math | 69.3% | — |
| GSM8K | — | 69.6% |
Knowledge GPT-4.5 leads
GPT-4.5: 32.5 (#211), Llama 2-70B: 7.4 (#310)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| GPQA Diamond | 68.7% | 26.3% |
| LMArena Expert | 1394 | 1039 |
| Humanity's Last Exam | 5.4% | — |
| Confabulations | 13.6% | — |
| ARC (AI2) Challenge | — | 78.3% |
| BoolQ | — | 88.6% |
| MMLU | — | 69.9% |
| OpenBookQA | — | 60.2% |
| TriviaQA | — | 87.6% |
Multimodal Not comparable
GPT-4.5: 37.6 (#71), Llama 2-70B: —
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| LMArena Vision | 1195 | — |
| VPCT | 45% | — |
Multilingual GPT-4.5 leads
GPT-4.5: 52.5 (#83), Llama 2-70B: 27.7 (#274)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 1413 | 1045 |
| LMArena Chinese | 1421 | 995 |
| LMArena French | 1418 | 1090 |
| LMArena German | 1457 | 1041 |
| LMArena Japanese | 1416 | 927 |
| LMArena Korean | 1392 | 964 |
| LMArena Russian | 1419 | 1083 |
| LMArena Spanish | — | 1143 |
Instruction Following GPT-4.5 leads
GPT-4.5: 72.6 (#134), Llama 2-70B: 54.9 (#278)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1404 | 1071 |
| LiveBench Instruction Following | 72.3% | — |
Long Context GPT-4.5 leads
GPT-4.5: 40.4 (#155), Llama 2-70B: 32.3 (#270)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1406 | 1062 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-4.5 leads
GPT-4.5: 56.9 (#134), Llama 2-70B: 32.3 (#279)
| Benchmark | GPT-4.5 | Llama 2-70B |
|---|---|---|
| LMArena Text | 1417 | 1115 |
| LMArena Creative Writing | 1394 | 1075 |
| LMArena Multi-Turn | 1444 | 1088 |
| Short-Story Creative Writing | 75.6% | — |
| EQ-Bench Creative Writing | 1258 | — |
| LiveBench Language | 61.5% | — |
Frequently asked questions
Is GPT-4.5 better than Llama 2-70B?
GPT-4.5 is the stronger model overall, scoring 37.2 to 24.4 on the Noometry Index.
Is GPT-4.5 or Llama 2-70B better for coding?
GPT-4.5 scores higher on coding benchmarks: 42.2 versus 31.4 in the Noometry coding category.
How many benchmarks do GPT-4.5 and Llama 2-70B share?
21 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Llama 2-70B has 35.