Model comparison
Phi 3 Mini 4k Instruct vs Qwen1.5-72B
Qwen1.5-72B is the stronger model overall, scoring 30.8 to 27.9 on the Noometry Index.
Last verified . 19 shared benchmarks.
Microsoft
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Phi 3 Mini 4k Instruct scores higher in 1 category and Qwen1.5-72B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Phi 3 Mini 4k Instruct leads 28.5 to 11.5.
Side by side
| Phi 3 Mini 4k Instruct | Qwen1.5-72B | |
|---|---|---|
| Provider | Microsoft | Alibaba (Qwen) |
| Noometry Index | 27.9 | 30.8 |
| Released | 2024-04-23 | 2024-02-04 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 35 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen1.5-72B leads
Phi 3 Mini 4k Instruct: 26.6 (#323), Qwen1.5-72B: 31.9 (#277)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Coding | 1093 | 1165 |
| HumanEval+ | 59.1% | 59.1% |
| MBPP+ | 54.2% | 61.6% |
| BigCodeBench Instruct | — | 33.2% |
| LiveBench Coding | 15.5% | — |
| BigCodeBench Complete | — | 40.3% |
Reasoning Qwen1.5-72B leads
Phi 3 Mini 4k Instruct: 14.1 (#328), Qwen1.5-72B: 22.2 (#203)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Hard Prompts | 1072 | 1148 |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | 26.8% | — |
| LiveBench Data Analysis | 34.7% | — |
| Adversarial NLI | 52.8% | — |
| BIG-Bench Hard | 71.7% | — |
| HellaSwag | 76.7% | — |
| LiveBench | 22.4% | — |
| WinoGrande | 70.8% | — |
Math Qwen1.5-72B leads
Phi 3 Mini 4k Instruct: 26.6 (#257), Qwen1.5-72B: 33.2 (#205)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Math | 1111 | 1164 |
| LiveBench Math | 15.7% | — |
Knowledge Phi 3 Mini 4k Instruct leads
Phi 3 Mini 4k Instruct: 28.5 (#246), Qwen1.5-72B: 11.5 (#300)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Expert | 1045 | 1136 |
| GPQA Diamond | — | 28.8% |
| ARC (AI2) Challenge | 84.9% | — |
| MMLU | 68.8% | — |
| OpenBookQA | 88% | — |
| TriviaQA | 64% | — |
Multilingual Qwen1.5-72B leads
Phi 3 Mini 4k Instruct: 26.3 (#280), Qwen1.5-72B: 33.2 (#253)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Non-English | 1021 | 1135 |
| LMArena Chinese | 1021 | 1186 |
| LMArena French | 1076 | 1159 |
| LMArena German | 1044 | 1084 |
| LMArena Japanese | 935 | 1061 |
| LMArena Korean | 905 | 1050 |
| LMArena Russian | 1022 | 1104 |
| LMArena Spanish | 1085 | 1110 |
Instruction Following Qwen1.5-72B leads
Phi 3 Mini 4k Instruct: 47.7 (#303), Qwen1.5-72B: 59.3 (#256)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Instruction Following | 1053 | 1141 |
| LiveBench Instruction Following | 39.1% | — |
Long Context Qwen1.5-72B leads
Phi 3 Mini 4k Instruct: 31.7 (#276), Qwen1.5-72B: 35.1 (#243)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Longer Query | 1044 | 1157 |
Writing & Preference Qwen1.5-72B leads
Phi 3 Mini 4k Instruct: 27.6 (#300), Qwen1.5-72B: 37.3 (#258)
| Benchmark | Phi 3 Mini 4k Instruct | Qwen1.5-72B |
|---|---|---|
| LMArena Text | 1073 | 1166 |
| LMArena Creative Writing | 1037 | 1137 |
| LMArena Multi-Turn | 1018 | 1160 |
| LiveBench Language | 9.2% | — |
Frequently asked questions
Is Phi 3 Mini 4k Instruct better than Qwen1.5-72B?
Qwen1.5-72B is the stronger model overall, scoring 30.8 to 27.9 on the Noometry Index.
Is Phi 3 Mini 4k Instruct or Qwen1.5-72B better for coding?
Qwen1.5-72B scores higher on coding benchmarks: 31.9 versus 26.6 in the Noometry coding category.
How many benchmarks do Phi 3 Mini 4k Instruct and Qwen1.5-72B share?
19 benchmarks have published results for both models. Phi 3 Mini 4k Instruct has 35 scored results on Noometry and Qwen1.5-72B has 22.