Model comparison
GPT-4 vs Llama 3-70B
GPT-4 and Llama 3-70B score almost the same on the Noometry Index (29.1 vs 28.8), so choose on price, context window or the category you care about most.
Last verified . 28 shared benchmarks.
MetaSummary
- They share 28 benchmarks with published results for both. GPT-4 scores higher in 3 categories and Llama 3-70B in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Llama 3-70B leads 42.8 to 34.9.
- The biggest single-benchmark swing is DTBench: 62.7% for GPT-4 and 54.2% for Llama 3-70B.
- Llama 3-70B has downloadable open weights; the other is API-only.
Side by side
| GPT-4 | Llama 3-70B | |
|---|---|---|
| Provider | OpenAI | Meta |
| Noometry Index | 29.1 | 28.8 |
| Released | 2023-03-14 | 2024-04-18 |
| Weights | Proprietary | Open |
| Context window | 8K | — |
| Max output | 8K | — |
| Input $ / M tokens | $30 | — |
| Output $ / M tokens | $60 | — |
| Results tracked | 38 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3-70B leads
GPT-4: 31.6 (#283), Llama 3-70B: 35.8 (#218)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| BigCodeBench Instruct | 46% | 43.6% |
| LMArena Coding | 1254 | 1206 |
| BigCodeBench Complete | 57.2% | 54.5% |
| HumanEval+ | 79.3% | 72% |
| WeirdML | 12.4% | — |
| MBPP+ | — | 69% |
Agentic & Tool Use Not comparable
GPT-4: —, Llama 3-70B: 21.1 (#139)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| Cybench | — | 5% |
| METR Time Horizons | 36.1% | — |
Reasoning Too close to call
GPT-4: 17.8 (#289), Llama 3-70B: 18.0 (#288)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1195 |
| DTBench | 62.7% | 54.2% |
| Epoch Capabilities Index | 125.89 | 122.93 |
| ForecastBench | 57.8 | 57.1 |
| WinoGrande | 87.5% | 83.5% |
| Kagi LLM Benchmark | — | 35.1% |
| Chess Puzzles | 4% | — |
| Mystery Game Puzzles | 12% | — |
| LMCA | 17.1% | — |
| BIG-Bench Hard | 75.1% | — |
| HellaSwag | 95.3% | — |
Math Llama 3-70B leads
GPT-4: 10.8 (#309), Llama 3-70B: 12.8 (#305)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 4.3% |
| LMArena Math | 1269 | 1218 |
| MATH Level 5 | 23% | 22.6% |
| GSM8K | 92% | — |
Knowledge Llama 3-70B leads
GPT-4: 18.4 (#282), Llama 3-70B: 20.8 (#277)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| GPQA Diamond | 35.7% | 40.6% |
| LMArena Expert | 1211 | 1149 |
| MMLU | 86.4% | 79.3% |
| TriviaQA | 84.8% | — |
Multilingual GPT-4 leads
GPT-4: 40.6 (#215), Llama 3-70B: 33.6 (#251)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| LMArena Non-English | 1246 | 1142 |
| LMArena Chinese | 1242 | 1114 |
| LMArena French | 1283 | 1232 |
| LMArena German | 1251 | 1169 |
| LMArena Japanese | 1209 | 1017 |
| LMArena Korean | 1184 | 1017 |
| LMArena Russian | 1251 | 1159 |
| LMArena Spanish | 1261 | 1241 |
Instruction Following GPT-4 leads
GPT-4: 65.3 (#222), Llama 3-70B: 62.5 (#238)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| LMArena Instruction Following | 1241 | 1194 |
Long Context GPT-4 leads
GPT-4: 37.7 (#212), Llama 3-70B: 35.6 (#240)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| LMArena Longer Query | 1244 | 1174 |
Writing & Preference Llama 3-70B leads
GPT-4: 34.9 (#268), Llama 3-70B: 42.8 (#231)
| Benchmark | GPT-4 | Llama 3-70B |
|---|---|---|
| LMArena Text | 1263 | 1221 |
| LMArena Creative Writing | 1244 | 1210 |
| LMArena Multi-Turn | 1257 | 1223 |
| EQ-Bench Creative Writing | 752 | — |
Frequently asked questions
Is GPT-4 better than Llama 3-70B?
GPT-4 and Llama 3-70B score almost the same on the Noometry Index (29.1 vs 28.8), so choose on price, context window or the category you care about most.
Is GPT-4 or Llama 3-70B better for coding?
Llama 3-70B scores higher on coding benchmarks: 35.8 versus 31.6 in the Noometry coding category.
How many benchmarks do GPT-4 and Llama 3-70B share?
28 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Llama 3-70B has 31.