Model comparison
Gemma 3 27B vs Phi-4
Gemma 3 27B and Phi-4 score almost the same on the Noometry Index (30.8 vs 31.2), so choose on price, context window or the category you care about most.
Last verified . 33 shared benchmarks.
Google
MicrosoftSummary
- They share 33 benchmarks with published results for both. Gemma 3 27B scores higher in 5 categories and Phi-4 in 4 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 3 27B leads 52.5 to 40.5.
- The biggest single-benchmark swing is Short-Story Creative Writing: 79.9% for Gemma 3 27B and 62.6% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.08 / $0.16 for Gemma 3 27B.
- Gemma 3 27B accepts more context: 131K tokens versus 128K.
Side by side
| Gemma 3 27B | Phi-4 | |
|---|---|---|
| Provider | Microsoft | |
| Noometry Index | 30.8 | 31.2 |
| Released | 2025-03-11 | 2024-12-11 |
| Weights | Open | Open |
| Context window | 131K | 128K |
| Max output | 8K | 4K |
| Input $ / M tokens | $0.08 | $0.07 |
| Output $ / M tokens | $0.16 | $0.14 |
| Results tracked | 43 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Phi-4 leads
Gemma 3 27B: 22.5 (#334), Phi-4: 34.4 (#239)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| LiveBench Coding | 39.9% | 30.7% |
| LMArena Coding | 1322 | 1231 |
| Aider Polyglot | 4.9% | — |
| SciCode | 21.2% | — |
| BigCodeBench Instruct | — | 45.5% |
| BigCodeBench Complete | — | 55.4% |
Agentic & Tool Use Gemma 3 27B leads
Gemma 3 27B: 25.1 (#110), Phi-4: 22.8 (#128)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | 28.8% |
| BALROG | — | 11.6% |
Reasoning Too close to call
Gemma 3 27B: 16.7 (#301), Phi-4: 17.7 (#291)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| Chess Puzzles | 0% | 1% |
| LiveBench Reasoning | 43.8% | 47.8% |
| LMArena Hard Prompts | 1340 | 1220 |
| LiveBench Data Analysis | 51.5% | 45.2% |
| Epoch Capabilities Index | 130.04 | 130.42 |
| LiveBench | 50% | 41.6% |
| Kagi LLM Benchmark | 40.4% | — |
| CritPt | 0% | — |
| DTBench | 52.5% | — |
| LMCA | 12.3% | — |
Math Gemma 3 27B leads
Gemma 3 27B: 25.9 (#265), Phi-4: 20.8 (#285)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 13.8% |
| LiveBench Math | 55.4% | 42% |
| LMArena Math | 1312 | 1246 |
| MATH Level 5 | 74% | 64.9% |
Knowledge Phi-4 leads
Gemma 3 27B: 25.5 (#261), Phi-4: 32.6 (#209)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| GPQA Diamond | 47.7% | 56.1% |
| Confabulations | 40.3% | 29.4% |
| Vectara Hallucination Rate | 7.4% | 3.7% |
| LMArena Expert | 1304 | 1203 |
| MMLU | — | 84.8% |
Multimodal Not comparable
Gemma 3 27B: 32.6 (#100), Phi-4: —
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| LMArena Vision | 1164 | — |
| GeoBench | 52% | — |
Multilingual Gemma 3 27B leads
Gemma 3 27B: 46.9 (#155), Phi-4: 37.2 (#237)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| LMArena Non-English | 1334 | 1197 |
| LMArena Chinese | 1346 | 1212 |
| LMArena French | 1368 | 1224 |
| LMArena German | 1362 | 1222 |
| LMArena Japanese | 1287 | 1158 |
| LMArena Korean | 1308 | 1151 |
| LMArena Russian | 1349 | 1209 |
| LMArena Spanish | 1349 | 1234 |
Instruction Following Gemma 3 27B leads
Gemma 3 27B: 70.6 (#160), Phi-4: 60.4 (#251)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| LiveBench Instruction Following | 74.9% | 58.4% |
| LMArena Instruction Following | 1321 | 1201 |
Long Context Phi-4 leads
Gemma 3 27B: 27.6 (#293), Phi-4: 36.9 (#226)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1333 | 1217 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Gemma 3 27B leads
Gemma 3 27B: 52.5 (#168), Phi-4: 40.5 (#244)
| Benchmark | Gemma 3 27B | Phi-4 |
|---|---|---|
| LMArena Text | 1358 | 1217 |
| LMArena Creative Writing | 1346 | 1182 |
| Short-Story Creative Writing | 79.9% | 62.6% |
| LMArena Multi-Turn | 1345 | 1206 |
| LiveBench Language | 34.6% | 25.6% |
| EQ-Bench Creative Writing | 1266 | — |
Frequently asked questions
Is Gemma 3 27B better than Phi-4?
Gemma 3 27B and Phi-4 score almost the same on the Noometry Index (30.8 vs 31.2), so choose on price, context window or the category you care about most.
Which is cheaper, Gemma 3 27B or Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Gemma 3 27B lists at $0.08 and $0.16.
Is Gemma 3 27B or Phi-4 better for coding?
Phi-4 scores higher on coding benchmarks: 34.4 versus 22.5 in the Noometry coding category.
Which has the bigger context window?
Gemma 3 27B does, with 131K tokens against 128K.
How many benchmarks do Gemma 3 27B and Phi-4 share?
33 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Phi-4 has 37.