Model comparison
Claude 2.1 vs GPT-4
GPT-4 is the stronger model overall, scoring 29.1 to 25.2 on the Noometry Index.
Last verified . 7 shared benchmarks.
Summary
- They share 7 benchmarks with published results for both. Claude 2.1 scores higher in 1 category and GPT-4 in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in coding, where GPT-4 leads 31.6 to 26.2.
- The biggest single-benchmark swing is DTBench: 51% for Claude 2.1 and 62.7% for GPT-4.
Side by side
| Claude 2.1 | GPT-4 | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 25.2 | 29.1 |
| Released | 2023-11-21 | 2023-03-14 |
| Weights | Proprietary | Proprietary |
| Context window | — | 8K |
| Max output | — | 8K |
| Input $ / M tokens | — | $30 |
| Output $ / M tokens | — | $60 |
| Results tracked | 7 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4 leads
Claude 2.1: 26.2 (#327), GPT-4: 31.6 (#283)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| WeirdML | 7.1% | 12.4% |
| BigCodeBench Instruct | — | 46% |
| LMArena Coding | — | 1254 |
| BigCodeBench Complete | — | 57.2% |
| HumanEval+ | — | 79.3% |
Agentic & Tool Use Not comparable
Claude 2.1: —, GPT-4: —
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| METR Time Horizons | — | 36.1% |
Reasoning Claude 2.1 leads
Claude 2.1: 21.4 (#221), GPT-4: 17.8 (#289)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| DTBench | 51% | 62.7% |
| Epoch Capabilities Index | 119.27 | 125.89 |
| ForecastBench | 54.2 | 57.8 |
| Chess Puzzles | — | 4% |
| LMArena Hard Prompts | — | 1241 |
| Mystery Game Puzzles | — | 12% |
| LMCA | — | 17.1% |
| BIG-Bench Hard | — | 75.1% |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |
Math Too close to call
Claude 2.1: 10.2 (#315), GPT-4: 10.8 (#309)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.9% | 1.1% |
| LMArena Math | — | 1269 |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |
Knowledge GPT-4 leads
Claude 2.1: 15.4 (#292), GPT-4: 18.4 (#282)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| GPQA Diamond | 33% | 35.7% |
| MMLU | 73.5% | 86.4% |
| LMArena Expert | — | 1211 |
| TriviaQA | — | 84.8% |
Multilingual Not comparable
Claude 2.1: —, GPT-4: 40.6 (#215)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| LMArena Non-English | — | 1246 |
| LMArena Chinese | — | 1242 |
| LMArena French | — | 1283 |
| LMArena German | — | 1251 |
| LMArena Japanese | — | 1209 |
| LMArena Korean | — | 1184 |
| LMArena Russian | — | 1251 |
| LMArena Spanish | — | 1261 |
Instruction Following Not comparable
Claude 2.1: —, GPT-4: 65.3 (#222)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| LMArena Instruction Following | — | 1241 |
Long Context Not comparable
Claude 2.1: —, GPT-4: 37.7 (#212)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| LMArena Longer Query | — | 1244 |
Writing & Preference Not comparable
Claude 2.1: —, GPT-4: 34.9 (#268)
| Benchmark | Claude 2.1 | GPT-4 |
|---|---|---|
| LMArena Text | — | 1263 |
| LMArena Creative Writing | — | 1244 |
| EQ-Bench Creative Writing | — | 752 |
| LMArena Multi-Turn | — | 1257 |
Frequently asked questions
Is Claude 2.1 better than GPT-4?
GPT-4 is the stronger model overall, scoring 29.1 to 25.2 on the Noometry Index.
Is Claude 2.1 or GPT-4 better for coding?
GPT-4 scores higher on coding benchmarks: 31.6 versus 26.2 in the Noometry coding category.
How many benchmarks do Claude 2.1 and GPT-4 share?
7 benchmarks have published results for both models. Claude 2.1 has 7 scored results on Noometry and GPT-4 has 38.