Model comparison
DeepSeek V4.1 Flash vs Phi-4
DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 31.2 on the Noometry Index. Phi-4 costs 3.0× less per token, which makes it the better buy when DeepSeek V4.1 Flash's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
DeepSeek
MicrosoftSummary
- They share 20 benchmarks with published results for both. DeepSeek V4.1 Flash scores higher in 9 categories and Phi-4 in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4.1 Flash leads 66.7 to 20.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 98.3% for DeepSeek V4.1 Flash and 13.8% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.15 / $0.60 for DeepSeek V4.1 Flash.
- DeepSeek V4.1 Flash accepts more context: 1M tokens versus 128K.
Side by side
| DeepSeek V4.1 Flash | Phi-4 | |
|---|---|---|
| Provider | DeepSeek | Microsoft |
| Noometry Index | 52.8 | 31.2 |
| Released | 2026-09-09 | 2024-12-11 |
| Weights | Open | Open |
| Context window | 1M | 128K |
| Max output | 393K | 4K |
| Input $ / M tokens | $0.15 | $0.07 |
| Output $ / M tokens | $0.60 | $0.14 |
| Results tracked | 37 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 52.9 (#32), Phi-4: 34.4 (#239)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| LMArena Coding | 1506 | 1231 |
| LMArena WebDev | 1619 | — |
| SciCode | 51.9% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
| ALE-Bench | 1,092 | — |
Agentic & Tool Use DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 31.2 (#69), Phi-4: 22.8 (#128)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| APEX-Agents | 39.5% | — |
| Berkeley Function Calling Leaderboard | — | 28.8% |
| BALROG | — | 11.6% |
| GDP.pdf | 19.8% | — |
Reasoning DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 50.2 (#36), Phi-4: 17.7 (#291)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| LMArena Hard Prompts | 1483 | 1220 |
| Epoch Capabilities Index | 154.9 | 130.42 |
| NYT Connections (extended) | 89.6% | — |
| CritPt | 14.3% | — |
| Chess Puzzles | — | 1% |
| LiveBench Reasoning | — | 47.8% |
| Mystery Game Puzzles | 43% | — |
| DTBench | 89.9% | — |
| LiveBench Data Analysis | — | 45.2% |
| LMCA | 47% | — |
| Surface Evolver Bench | 46.3% | — |
| LiveBench | — | 41.6% |
Math DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 66.7 (#25), Phi-4: 20.8 (#285)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.3% | 13.8% |
| LMArena Math | 1477 | 1246 |
| FrontierMath (Tiers 1-3) | 67.4% | — |
| FrontierMath Tier 4 | 26.8% | — |
| ProofBench | 54% | — |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
Knowledge DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 57.9 (#38), Phi-4: 32.6 (#209)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| GPQA Diamond | 89.8% | 56.1% |
| LMArena Expert | 1506 | 1203 |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| MMLU | — | 84.8% |
Multimodal Not comparable
DeepSeek V4.1 Flash: 39.1 (#61), Phi-4: —
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| LMArena Vision | 1277 | — |
| Furniture Assembly | 34.2% | — |
Multilingual DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 55.0 (#35), Phi-4: 37.2 (#237)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| LMArena Non-English | 1448 | 1197 |
| LMArena Chinese | 1497 | 1212 |
| LMArena French | 1452 | 1224 |
| LMArena German | 1484 | 1222 |
| LMArena Japanese | 1412 | 1158 |
| LMArena Korean | 1452 | 1151 |
| LMArena Russian | 1471 | 1209 |
| LMArena Spanish | 1459 | 1234 |
Instruction Following DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 77.3 (#26), Phi-4: 60.4 (#251)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1474 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
Long Context DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 45.2 (#47), Phi-4: 36.9 (#226)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1475 | 1217 |
Writing & Preference DeepSeek V4.1 Flash leads
DeepSeek V4.1 Flash: 65.4 (#48), Phi-4: 40.5 (#244)
| Benchmark | DeepSeek V4.1 Flash | Phi-4 |
|---|---|---|
| LMArena Text | 1462 | 1217 |
| LMArena Creative Writing | 1435 | 1182 |
| LMArena Multi-Turn | 1457 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| EQ-Bench Creative Writing | 1540 | — |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is DeepSeek V4.1 Flash better than Phi-4?
DeepSeek V4.1 Flash is the stronger model overall, scoring 52.8 to 31.2 on the Noometry Index. Phi-4 costs 3.0× less per token, which makes it the better buy when DeepSeek V4.1 Flash's lead doesn't matter for your workload.
Which is cheaper, DeepSeek V4.1 Flash or Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; DeepSeek V4.1 Flash lists at $0.15 and $0.60.
Is DeepSeek V4.1 Flash or Phi-4 better for coding?
DeepSeek V4.1 Flash scores higher on coding benchmarks: 52.9 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4.1 Flash does, with 1M tokens against 128K.
How many benchmarks do DeepSeek V4.1 Flash and Phi-4 share?
20 benchmarks have published results for both models. DeepSeek V4.1 Flash has 37 scored results on Noometry and Phi-4 has 37.