Model comparison
DeepSeek V4 Flash vs Phi 3 Mini 4k Instruct
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 27.4 on the Noometry Index.
Last verified . 18 shared benchmarks.
DeepSeek
MicrosoftSummary
- They share 18 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 8 categories and Phi 3 Mini 4k Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 13.6.
- The biggest single-benchmark swing is Chess Puzzles: 33% for DeepSeek V4 Flash and 0% for Phi 3 Mini 4k Instruct.
Side by side
| DeepSeek V4 Flash | Phi 3 Mini 4k Instruct | |
|---|---|---|
| Provider | DeepSeek | Microsoft |
| Noometry Index | 53.6 | 27.4 |
| Released | 2026-04-24 | 2024-04-23 |
| Weights | Open | Open |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.15 | — |
| Output $ / M tokens | $0.60 | — |
| Results tracked | 41 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Flash leads
DeepSeek V4 Flash: 47.6 (#56), Phi 3 Mini 4k Instruct: 26.1 (#319)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Coding | 1457 | 1093 |
| FrontierCode | 18.8% | — |
| LMArena WebDev | 1581 | — |
| SciCode | 49.9% | — |
| WeirdML | 63% | — |
| LiveBench Coding | — | 15.5% |
| ALE-Bench | 1,306 | — |
| HumanEval+ | — | 59.1% |
| MBPP+ | — | 54.2% |
Reasoning DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.7 (#25), Phi 3 Mini 4k Instruct: 13.6 (#326)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| Chess Puzzles | 33% | 0% |
| LMArena Hard Prompts | 1443 | 1072 |
| ARC-AGI-2 | 61.4% | — |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | 89% | — |
| CritPt | 16.6% | — |
| LiveBench Reasoning | — | 26.8% |
| Mystery Game Puzzles | 34% | — |
| DTBench | 90.9% | — |
| LiveBench Data Analysis | — | 34.7% |
| LMCA | 41.7% | — |
| Adversarial NLI | — | 52.8% |
| BIG-Bench Hard | — | 71.7% |
| Epoch Capabilities Index | 154.51 | — |
| HellaSwag | — | 76.7% |
| LiveBench | — | 22.4% |
| WinoGrande | — | 70.8% |
Math DeepSeek V4 Flash leads
DeepSeek V4 Flash: 60.0 (#31), Phi 3 Mini 4k Instruct: 24.5 (#258)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Math | 1427 | 1111 |
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| OTIS Mock AIME 2024-2025 | 94.4% | — |
| ProofBench | 56% | — |
| LiveBench Math | — | 15.7% |
Knowledge DeepSeek V4 Flash leads
DeepSeek V4 Flash: 55.5 (#43), Phi 3 Mini 4k Instruct: 27.9 (#241)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Expert | 1441 | 1045 |
| GPQA Diamond | 91% | — |
| SimpleQA Verified | 33.6% | — |
| ARC (AI2) Challenge | — | 84.9% |
| MMLU | — | 68.8% |
| OpenBookQA | — | 88% |
| TriviaQA | — | 64% |
Multilingual DeepSeek V4 Flash leads
DeepSeek V4 Flash: 53.2 (#71), Phi 3 Mini 4k Instruct: 26.5 (#275)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Non-English | 1420 | 1021 |
| LMArena Chinese | 1467 | 1021 |
| LMArena French | 1438 | 1075 |
| LMArena German | 1418 | 1044 |
| LMArena Japanese | 1403 | 935 |
| LMArena Korean | 1383 | 905 |
| LMArena Russian | 1428 | 1022 |
| LMArena Spanish | 1437 | 1084 |
Instruction Following DeepSeek V4 Flash leads
DeepSeek V4 Flash: 75.1 (#76), Phi 3 Mini 4k Instruct: 48.0 (#298)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Instruction Following | 1421 | 1053 |
| LiveBench Instruction Following | — | 39.1% |
Long Context DeepSeek V4 Flash leads
DeepSeek V4 Flash: 43.8 (#81), Phi 3 Mini 4k Instruct: 31.7 (#271)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Longer Query | 1435 | 1044 |
Writing & Preference DeepSeek V4 Flash leads
DeepSeek V4 Flash: 64.0 (#59), Phi 3 Mini 4k Instruct: 27.7 (#295)
| Benchmark | DeepSeek V4 Flash | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Text | 1432 | 1073 |
| LMArena Creative Writing | 1403 | 1037 |
| LMArena Multi-Turn | 1449 | 1018 |
| EQ-Bench Creative Writing | 1559 | — |
| LiveBench Language | — | 9.2% |
Frequently asked questions
Is DeepSeek V4 Flash better than Phi 3 Mini 4k Instruct?
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 27.4 on the Noometry Index.
Is DeepSeek V4 Flash or Phi 3 Mini 4k Instruct better for coding?
DeepSeek V4 Flash scores higher on coding benchmarks: 47.6 versus 26.1 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Flash and Phi 3 Mini 4k Instruct share?
18 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Phi 3 Mini 4k Instruct has 35.