Model comparison
GPT-4 vs Qwen2.5 7B Instruct
GPT-4 and Qwen2.5 7B Instruct score almost the same on the Noometry Index (29.1 vs 29.0), so choose on price, context window or the category you care about most.
Last verified . 9 shared benchmarks.
Alibaba (Qwen)Summary
- They share 9 benchmarks with published results for both. GPT-4 scores higher in 3 categories and Qwen2.5 7B Instruct in 3 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen2.5 7B Instruct leads 48.8 to 34.9.
- The biggest single-benchmark swing is DTBench: 62.7% for GPT-4 and 47.7% for Qwen2.5 7B Instruct.
- Qwen2.5 7B Instruct is cheaper at $0.17 / $0.70 per million input/output tokens, against $30 / $60 for GPT-4.
- Qwen2.5 7B Instruct accepts more context: 131K tokens versus 8K.
- Qwen2.5 7B Instruct has downloadable open weights; the other is API-only.
Side by side
| GPT-4 | Qwen2.5 7B Instruct | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 29.1 | 29.0 |
| Released | 2023-03-14 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 8K | 131K |
| Max output | 8K | 8K |
| Input $ / M tokens | $30 | $0.17 |
| Output $ / M tokens | $60 | $0.70 |
| Results tracked | 38 | 15 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen2.5 7B Instruct leads
GPT-4: 31.6 (#283), Qwen2.5 7B Instruct: 36.5 (#208)
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| BigCodeBench Instruct | 46% | 37.6% |
| BigCodeBench Complete | 57.2% | 46.1% |
| WeirdML | 12.4% | — |
| LMArena Coding | 1254 | — |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, Qwen2.5 7B Instruct: 23.8 (#124)
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| BALROG | — | 7.8% |
| METR Time Horizons | 36.1% | — |
Reasoning GPT-4 leads
GPT-4: 17.8 (#289), Qwen2.5 7B Instruct: 14.8 (#322)
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| Chess Puzzles | 4% | 0% |
| DTBench | 62.7% | 47.7% |
| LMCA | 17.1% | 6.4% |
| Epoch Capabilities Index | 125.89 | 118.51 |
| LMArena Hard Prompts | 1241 | — |
| Mystery Game Puzzles | 12% | — |
| BIG-Bench Hard | 75.1% | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |
Math Qwen2.5 7B Instruct leads
GPT-4: 10.8 (#309), Qwen2.5 7B Instruct: 12.6 (#306)
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 2.5% |
| Omni-MATH | — | 29.4% |
| LMArena Math | 1269 | — |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |
Knowledge GPT-4 leads
GPT-4: 18.4 (#282), Qwen2.5 7B Instruct: 17.0 (#286)
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| GPQA Diamond | 35.7% | 35.5% |
| MMLU | 86.4% | 72.9% |
| MMLU-Pro | — | 53.9% |
| GPQA (HELM) | — | 34.1% |
| LMArena Expert | 1211 | — |
| TriviaQA | 84.8% | — |
Multilingual Not comparable
GPT-4: 40.6 (#215), Qwen2.5 7B Instruct: —
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| 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 GPT-4 leads
GPT-4: 65.3 (#222), Qwen2.5 7B Instruct: 63.2 (#231)
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| IFEval | — | 74.1% |
| LMArena Instruction Following | 1241 | — |
Long Context Not comparable
GPT-4: 37.7 (#212), Qwen2.5 7B Instruct: —
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Longer Query | 1244 | — |
Writing & Preference Qwen2.5 7B Instruct leads
GPT-4: 34.9 (#268), Qwen2.5 7B Instruct: 48.8 (#195)
| Benchmark | GPT-4 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Text | 1263 | — |
| LMArena Creative Writing | 1244 | — |
| EQ-Bench Creative Writing | 752 | — |
| WildBench | — | 73.1% |
| LMArena Multi-Turn | 1257 | — |
Frequently asked questions
Is GPT-4 better than Qwen2.5 7B Instruct?
GPT-4 and Qwen2.5 7B Instruct score almost the same on the Noometry Index (29.1 vs 29.0), so choose on price, context window or the category you care about most.
Which is cheaper, GPT-4 or Qwen2.5 7B Instruct?
Qwen2.5 7B Instruct is cheaper. It lists at $0.17 per million input tokens and $0.70 per million output tokens; GPT-4 lists at $30 and $60.
Is GPT-4 or Qwen2.5 7B Instruct better for coding?
Qwen2.5 7B Instruct scores higher on coding benchmarks: 36.5 versus 31.6 in the Noometry coding category.
Which has the bigger context window?
Qwen2.5 7B Instruct does, with 131K tokens against 8K.
How many benchmarks do GPT-4 and Qwen2.5 7B Instruct share?
9 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Qwen2.5 7B Instruct has 15.