Model comparison
Phi-4 vs Qwen3.5-Flash
Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 31.2 on the Noometry Index. Phi-4 costs 2.0× less per token, which makes it the better buy when Qwen3.5-Flash's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
Microsoft
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. Phi-4 scores higher in 1 category and Qwen3.5-Flash in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.5-Flash leads 57.9 to 40.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 13.8% for Phi-4 and 84.4% for Qwen3.5-Flash.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.10 / $0.40 for Qwen3.5-Flash.
- Qwen3.5-Flash accepts more context: 1M tokens versus 128K.
- Phi-4 has downloadable open weights; the other is API-only.
Side by side
| Phi-4 | Qwen3.5-Flash | |
|---|---|---|
| Provider | Microsoft | Alibaba (Qwen) |
| Noometry Index | 31.2 | 42.5 |
| Released | 2024-12-11 | 2026-02-23 |
| Weights | Open | Proprietary |
| Context window | 128K | 1M |
| Max output | 4K | 66K |
| Input $ / M tokens | $0.07 | $0.10 |
| Output $ / M tokens | $0.14 | $0.40 |
| Results tracked | 37 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Phi-4: 34.4 (#239), Qwen3.5-Flash: 34.2 (#242)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| LMArena Coding | 1231 | 1412 |
| LMArena WebDev | — | 1244 |
| BigCodeBench Instruct | 45.5% | — |
| LiveBench Coding | 30.7% | — |
| BigCodeBench Complete | 55.4% | — |
| ALE-Bench | — | 221.8 |
Agentic & Tool Use Not comparable
Phi-4: 22.8 (#128), Qwen3.5-Flash: —
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| Berkeley Function Calling Leaderboard | 28.8% | — |
| BALROG | 11.6% | — |
| Vending-Bench 2 | — | 462.69 |
Reasoning Qwen3.5-Flash leads
Phi-4: 17.7 (#291), Qwen3.5-Flash: 33.7 (#72)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| Chess Puzzles | 1% | 21% |
| LMArena Hard Prompts | 1220 | 1403 |
| Epoch Capabilities Index | 130.42 | 143.98 |
| LiveBench Reasoning | 47.8% | — |
| Mystery Game Puzzles | — | 20% |
| DTBench | — | 82.9% |
| LiveBench Data Analysis | 45.2% | — |
| LMCA | — | 29.1% |
| LiveBench | 41.6% | — |
Math Qwen3.5-Flash leads
Phi-4: 20.8 (#285), Qwen3.5-Flash: 37.4 (#158)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 13.8% | 84.4% |
| LMArena Math | 1246 | 1407 |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| LiveBench Math | 42% | — |
| MATH Level 5 | 64.9% | — |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3.5-Flash leads
Phi-4: 32.6 (#209), Qwen3.5-Flash: 43.2 (#93)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| GPQA Diamond | 56.1% | 82.3% |
| Vectara Hallucination Rate | 3.7% | 10.5% |
| LMArena Expert | 1203 | 1407 |
| SimpleQA Verified | — | 20.3% |
| Confabulations | 29.4% | — |
| MMLU | 84.8% | — |
Multilingual Qwen3.5-Flash leads
Phi-4: 37.2 (#237), Qwen3.5-Flash: 50.5 (#121)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | 1197 | 1385 |
| LMArena Chinese | 1212 | 1446 |
| LMArena French | 1224 | 1412 |
| LMArena German | 1222 | 1390 |
| LMArena Japanese | 1158 | 1368 |
| LMArena Korean | 1151 | 1344 |
| LMArena Russian | 1209 | 1379 |
| LMArena Spanish | 1234 | 1400 |
Instruction Following Qwen3.5-Flash leads
Phi-4: 60.4 (#251), Qwen3.5-Flash: 72.6 (#139)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | 1201 | 1374 |
| LiveBench Instruction Following | 58.4% | — |
Long Context Qwen3.5-Flash leads
Phi-4: 36.9 (#226), Qwen3.5-Flash: 42.4 (#124)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | 1217 | 1392 |
Writing & Preference Qwen3.5-Flash leads
Phi-4: 40.5 (#244), Qwen3.5-Flash: 57.9 (#122)
| Benchmark | Phi-4 | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | 1217 | 1397 |
| LMArena Creative Writing | 1182 | 1343 |
| LMArena Multi-Turn | 1206 | 1393 |
| Short-Story Creative Writing | 62.6% | — |
| LiveBench Language | 25.6% | — |
Frequently asked questions
Is Phi-4 better than Qwen3.5-Flash?
Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 31.2 on the Noometry Index. Phi-4 costs 2.0× less per token, which makes it the better buy when Qwen3.5-Flash's lead doesn't matter for your workload.
Which is cheaper, Phi-4 or Qwen3.5-Flash?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Qwen3.5-Flash lists at $0.10 and $0.40.
Is Phi-4 or Qwen3.5-Flash better for coding?
They score almost the same on coding (34.4 vs 34.2); test both on your own repository before choosing.
Which has the bigger context window?
Qwen3.5-Flash does, with 1M tokens against 128K.
How many benchmarks do Phi-4 and Qwen3.5-Flash share?
22 benchmarks have published results for both models. Phi-4 has 37 scored results on Noometry and Qwen3.5-Flash has 32.