Model comparison
GPT-4 vs Qwen1.5-110B
Qwen1.5-110B is the stronger model overall, scoring 34.2 to 29.1 on the Noometry Index.
Last verified . 20 shared benchmarks.
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. GPT-4 scores higher in 3 categories and Qwen1.5-110B in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen1.5-110B leads 33.7 to 10.8.
- The biggest single-benchmark swing is BigCodeBench Complete: 57.2% for GPT-4 and 44.4% for Qwen1.5-110B.
- Qwen1.5-110B has downloadable open weights; the other is API-only.
Side by side
| GPT-4 | Qwen1.5-110B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 29.1 | 34.2 |
| Released | 2023-03-14 | 2024-04-25 |
| Weights | Proprietary | Open |
| Context window | 8K | — |
| Max output | 8K | — |
| Input $ / M tokens | $30 | — |
| Output $ / M tokens | $60 | — |
| Results tracked | 38 | 20 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen1.5-110B leads
GPT-4: 31.6 (#283), Qwen1.5-110B: 33.0 (#264)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| BigCodeBench Instruct | 46% | 35% |
| LMArena Coding | 1254 | 1184 |
| BigCodeBench Complete | 57.2% | 44.4% |
| WeirdML | 12.4% | — |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, Qwen1.5-110B: —
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| METR Time Horizons | 36.1% | — |
Reasoning Qwen1.5-110B leads
GPT-4: 17.8 (#289), Qwen1.5-110B: 22.7 (#189)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1168 |
| ForecastBench | 57.8 | 57.7 |
| Chess Puzzles | 4% | — |
| Mystery Game Puzzles | 12% | — |
| DTBench | 62.7% | — |
| LMCA | 17.1% | — |
| BIG-Bench Hard | 75.1% | — |
| Epoch Capabilities Index | 125.89 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |
Math Qwen1.5-110B leads
GPT-4: 10.8 (#309), Qwen1.5-110B: 33.7 (#201)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| LMArena Math | 1269 | 1185 |
| OTIS Mock AIME 2024-2025 | 1.1% | — |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |
Knowledge Qwen1.5-110B leads
GPT-4: 18.4 (#282), Qwen1.5-110B: 31.2 (#219)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| LMArena Expert | 1211 | 1144 |
| GPQA Diamond | 35.7% | — |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multilingual GPT-4 leads
GPT-4: 40.6 (#215), Qwen1.5-110B: 33.6 (#250)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| LMArena Non-English | 1246 | 1142 |
| LMArena Chinese | 1242 | 1206 |
| LMArena French | 1283 | 1151 |
| LMArena German | 1251 | 1123 |
| LMArena Japanese | 1209 | 1074 |
| LMArena Korean | 1184 | 1044 |
| LMArena Russian | 1251 | 1118 |
| LMArena Spanish | 1261 | 1142 |
Instruction Following GPT-4 leads
GPT-4: 65.3 (#222), Qwen1.5-110B: 60.3 (#252)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| LMArena Instruction Following | 1241 | 1158 |
Long Context GPT-4 leads
GPT-4: 37.7 (#212), Qwen1.5-110B: 35.1 (#242)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| LMArena Longer Query | 1244 | 1157 |
Writing & Preference Qwen1.5-110B leads
GPT-4: 34.9 (#268), Qwen1.5-110B: 38.0 (#255)
| Benchmark | GPT-4 | Qwen1.5-110B |
|---|---|---|
| LMArena Text | 1263 | 1175 |
| LMArena Creative Writing | 1244 | 1148 |
| LMArena Multi-Turn | 1257 | 1160 |
| EQ-Bench Creative Writing | 752 | — |
Frequently asked questions
Is GPT-4 better than Qwen1.5-110B?
Qwen1.5-110B is the stronger model overall, scoring 34.2 to 29.1 on the Noometry Index.
Is GPT-4 or Qwen1.5-110B better for coding?
Qwen1.5-110B scores higher on coding benchmarks: 33.0 versus 31.6 in the Noometry coding category.
How many benchmarks do GPT-4 and Qwen1.5-110B share?
20 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Qwen1.5-110B has 20.