Model comparison
Phi-4 vs Qwen3.7 Max
Qwen3.7 Max is the stronger model overall, scoring 51.5 to 31.2 on the Noometry Index. Phi-4 costs 43× less per token, which makes it the better buy when Qwen3.7 Max's lead doesn't matter for your workload.
Last verified . 16 shared benchmarks.
Microsoft
Alibaba (Qwen)Summary
- They share 16 benchmarks with published results for both. Phi-4 scores higher in 1 category and Qwen3.7 Max in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.7 Max leads 62.4 to 20.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 13.8% for Phi-4 and 95.6% for Qwen3.7 Max.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $2.50 / $7.50 for Qwen3.7 Max.
- Qwen3.7 Max accepts more context: 1M tokens versus 128K.
- Phi-4 has downloadable open weights; the other is API-only.
Side by side
| Phi-4 | Qwen3.7 Max | |
|---|---|---|
| Provider | Microsoft | Alibaba (Qwen) |
| Noometry Index | 31.2 | 51.5 |
| Released | 2024-12-11 | 2026-05-19 |
| Weights | Open | Proprietary |
| Context window | 128K | 1M |
| Max output | 4K | 131K |
| Input $ / M tokens | $0.07 | $2.50 |
| Output $ / M tokens | $0.14 | $7.50 |
| Results tracked | 37 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.7 Max leads
Phi-4: 34.4 (#239), Qwen3.7 Max: 50.4 (#45)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| LMArena Coding | 1231 | 1498 |
| SWE-bench Verified | — | 77.3% |
| LMArena WebDev | — | 1515 |
| SciCode | — | 48.8% |
| BigCodeBench Instruct | 45.5% | — |
| LiveBench Coding | 30.7% | — |
| BigCodeBench Complete | 55.4% | — |
| ALE-Bench | — | 1,189 |
Agentic & Tool Use Too close to call
Phi-4: 22.8 (#128), Qwen3.7 Max: 22.1 (#135)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| Berkeley Function Calling Leaderboard | 28.8% | — |
| BALROG | 11.6% | — |
| GBAEval | — | 0.4% |
Reasoning Qwen3.7 Max leads
Phi-4: 17.7 (#291), Qwen3.7 Max: 49.2 (#38)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| Chess Puzzles | 1% | 19% |
| LMArena Hard Prompts | 1220 | 1483 |
| Epoch Capabilities Index | 130.42 | 153.68 |
| SimpleBench | — | 70.4% |
| NYT Connections (extended) | — | 85.1% |
| CritPt | — | 13.4% |
| EBR-Bench | — | 9.5% |
| LiveBench Reasoning | 47.8% | — |
| Mystery Game Puzzles | — | 32% |
| DTBench | — | 92.3% |
| LiveBench Data Analysis | 45.2% | — |
| LMCA | — | 44% |
| LiveBench | 41.6% | — |
Math Qwen3.7 Max leads
Phi-4: 20.8 (#285), Qwen3.7 Max: 62.4 (#32)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 13.8% | 95.6% |
| LMArena Math | 1246 | 1490 |
| FrontierMath (Tiers 1-3) | — | 64.6% |
| FrontierMath Tier 4 | — | 34.1% |
| ProofBench | — | 26% |
| LiveBench Math | 42% | — |
| MATH Level 5 | 64.9% | — |
Knowledge Qwen3.7 Max leads
Phi-4: 32.6 (#209), Qwen3.7 Max: 61.6 (#28)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| GPQA Diamond | 56.1% | 90.9% |
| LMArena Expert | 1203 | 1488 |
| SimpleQA Verified | — | 55.8% |
| Confabulations | 29.4% | — |
| Vectara Hallucination Rate | 3.7% | — |
| MMLU | 84.8% | — |
Multilingual Qwen3.7 Max leads
Phi-4: 37.2 (#237), Qwen3.7 Max: 56.9 (#15)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| LMArena Non-English | 1197 | 1474 |
| LMArena Chinese | 1212 | 1530 |
| LMArena Russian | 1209 | 1484 |
| LMArena French | 1224 | — |
| LMArena German | 1222 | — |
| LMArena Japanese | 1158 | — |
| LMArena Korean | 1151 | — |
| LMArena Spanish | 1234 | — |
Instruction Following Qwen3.7 Max leads
Phi-4: 60.4 (#251), Qwen3.7 Max: 76.7 (#38)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| LMArena Instruction Following | 1201 | 1460 |
| LiveBench Instruction Following | 58.4% | — |
Long Context Qwen3.7 Max leads
Phi-4: 36.9 (#226), Qwen3.7 Max: 45.4 (#40)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| LMArena Longer Query | 1217 | 1482 |
Writing & Preference Qwen3.7 Max leads
Phi-4: 40.5 (#244), Qwen3.7 Max: 65.0 (#54)
| Benchmark | Phi-4 | Qwen3.7 Max |
|---|---|---|
| LMArena Text | 1217 | 1476 |
| LMArena Creative Writing | 1182 | 1449 |
| LMArena Multi-Turn | 1206 | 1481 |
| Short-Story Creative Writing | 62.6% | — |
| EQ-Bench 4 | — | 1110 |
| LiveBench Language | 25.6% | — |
Frequently asked questions
Is Phi-4 better than Qwen3.7 Max?
Qwen3.7 Max is the stronger model overall, scoring 51.5 to 31.2 on the Noometry Index. Phi-4 costs 43× less per token, which makes it the better buy when Qwen3.7 Max's lead doesn't matter for your workload.
Which is cheaper, Phi-4 or Qwen3.7 Max?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Qwen3.7 Max lists at $2.50 and $7.50.
Is Phi-4 or Qwen3.7 Max better for coding?
Qwen3.7 Max scores higher on coding benchmarks: 50.4 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
Qwen3.7 Max does, with 1M tokens against 128K.
How many benchmarks do Phi-4 and Qwen3.7 Max share?
16 benchmarks have published results for both models. Phi-4 has 37 scored results on Noometry and Qwen3.7 Max has 33.