Model comparison
Gemma 4 31B IT vs Phi 3 Mini 4k Instruct
Gemma 4 31B IT is the stronger model overall, scoring 43.5 to 27.9 on the Noometry Index.
Last verified . 15 shared benchmarks.
Google
MicrosoftSummary
- They share 15 benchmarks with published results for both. Gemma 4 31B IT 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 writing & preference, where Gemma 4 31B IT leads 60.5 to 27.6.
- The biggest single-benchmark swing is Chess Puzzles: 5% for Gemma 4 31B IT and 0% for Phi 3 Mini 4k Instruct.
Side by side
| Gemma 4 31B IT | Phi 3 Mini 4k Instruct | |
|---|---|---|
| Provider | Microsoft | |
| Noometry Index | 43.5 | 27.9 |
| Released | 2026-04-02 | 2024-04-23 |
| Weights | Open | Open |
| Context window | 262K | — |
| Max output | 33K | — |
| Input $ / M tokens | $0.09 | — |
| Output $ / M tokens | $0.34 | — |
| Results tracked | 35 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 4 31B IT leads
Gemma 4 31B IT: 42.3 (#108), Phi 3 Mini 4k Instruct: 26.6 (#323)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Coding | 1459 | 1093 |
| LMArena WebDev | 1366 | — |
| SciCode | 43.4% | — |
| WeirdML | 52.3% | — |
| LiveBench Coding | — | 15.5% |
| ALE-Bench | 925.5 | — |
| HumanEval+ | — | 59.1% |
| MBPP+ | — | 54.2% |
Reasoning Gemma 4 31B IT leads
Gemma 4 31B IT: 27.2 (#122), Phi 3 Mini 4k Instruct: 14.1 (#328)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| Chess Puzzles | 5% | 0% |
| LMArena Hard Prompts | 1448 | 1072 |
| Kagi LLM Benchmark | 63.5% | — |
| NYT Connections (extended) | 70.6% | — |
| CritPt | 1.4% | — |
| Thematic Generalization | 53% | — |
| LiveBench Reasoning | — | 26.8% |
| DTBench | 82.7% | — |
| LiveBench Data Analysis | — | 34.7% |
| LMCA | 39.3% | — |
| Surface Evolver Bench | 30.6% | — |
| Adversarial NLI | — | 52.8% |
| BIG-Bench Hard | — | 71.7% |
| Epoch Capabilities Index | 142.74 | — |
| HellaSwag | — | 76.7% |
| LiveBench | — | 22.4% |
| WinoGrande | — | 70.8% |
Math Gemma 4 31B IT leads
Gemma 4 31B IT: 43.2 (#81), Phi 3 Mini 4k Instruct: 26.6 (#257)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Math | 1465 | 1111 |
| OTIS Mock AIME 2024-2025 | 73.3% | — |
| LiveBench Math | — | 15.7% |
Knowledge Gemma 4 31B IT leads
Gemma 4 31B IT: 37.9 (#151), Phi 3 Mini 4k Instruct: 28.5 (#246)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Expert | 1465 | 1045 |
| GPQA Diamond | 75.8% | — |
| SimpleQA Verified | 10.4% | — |
| Vectara Hallucination Rate | 7.4% | — |
| ARC (AI2) Challenge | — | 84.9% |
| MMLU | — | 68.8% |
| OpenBookQA | — | 88% |
| TriviaQA | — | 64% |
Multimodal Not comparable
Gemma 4 31B IT: 41.6 (#34), Phi 3 Mini 4k Instruct: —
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Vision | 1277 | — |
| LMArena Document | 1425 | — |
Multilingual Gemma 4 31B IT leads
Gemma 4 31B IT: 53.8 (#57), Phi 3 Mini 4k Instruct: 26.3 (#280)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Non-English | 1431 | 1021 |
| LMArena Chinese | 1476 | 1021 |
| LMArena French | 1435 | 1076 |
| LMArena Russian | 1460 | 1022 |
| LMArena Spanish | 1444 | 1085 |
| LMArena German | — | 1044 |
| LMArena Japanese | — | 935 |
| LMArena Korean | — | 905 |
Instruction Following Gemma 4 31B IT leads
Gemma 4 31B IT: 75.5 (#61), Phi 3 Mini 4k Instruct: 47.7 (#303)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Instruction Following | 1433 | 1053 |
| LiveBench Instruction Following | — | 39.1% |
Long Context Gemma 4 31B IT leads
Gemma 4 31B IT: 44.2 (#71), Phi 3 Mini 4k Instruct: 31.7 (#276)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Longer Query | 1446 | 1044 |
Writing & Preference Gemma 4 31B IT leads
Gemma 4 31B IT: 60.5 (#96), Phi 3 Mini 4k Instruct: 27.6 (#300)
| Benchmark | Gemma 4 31B IT | Phi 3 Mini 4k Instruct |
|---|---|---|
| LMArena Text | 1443 | 1073 |
| LMArena Creative Writing | 1415 | 1037 |
| LMArena Multi-Turn | 1452 | 1018 |
| EQ-Bench Creative Writing | 1368 | — |
| EQ-Bench 4 | 1120 | — |
| LiveBench Language | — | 9.2% |
Frequently asked questions
Is Gemma 4 31B IT better than Phi 3 Mini 4k Instruct?
Gemma 4 31B IT is the stronger model overall, scoring 43.5 to 27.9 on the Noometry Index.
Is Gemma 4 31B IT or Phi 3 Mini 4k Instruct better for coding?
Gemma 4 31B IT scores higher on coding benchmarks: 42.3 versus 26.6 in the Noometry coding category.
How many benchmarks do Gemma 4 31B IT and Phi 3 Mini 4k Instruct share?
15 benchmarks have published results for both models. Gemma 4 31B IT has 35 scored results on Noometry and Phi 3 Mini 4k Instruct has 35.