Model comparison
Grok 4.1 vs Phi-4
Grok 4.1 is the stronger model overall, scoring 41.5 to 31.2 on the Noometry Index.
Last verified . 17 shared benchmarks.
MicrosoftSummary
- They share 17 benchmarks with published results for both. Grok 4.1 scores higher in 8 categories and Phi-4 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Grok 4.1 leads 62.4 to 40.5.
- Phi-4 has downloadable open weights; the other is API-only.
Side by side
| Grok 4.1 | Phi-4 | |
|---|---|---|
| Provider | xAI | Microsoft |
| Noometry Index | 41.5 | 31.2 |
| Released | 2025-11-17 | 2024-12-11 |
| Weights | Proprietary | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.07 |
| Output $ / M tokens | — | $0.14 |
| Results tracked | 19 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Grok 4.1: 33.7 (#253), Phi-4: 34.4 (#239)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Coding | 1445 | 1231 |
| LMArena WebDev | 1214 | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
Agentic & Tool Use Grok 4.1 leads
Grok 4.1: 34.1 (#49), Phi-4: 22.8 (#128)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.8% |
| Cybench | 39% | — |
| BALROG | — | 11.6% |
Reasoning Grok 4.1 leads
Grok 4.1: 29.5 (#91), Phi-4: 17.7 (#291)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Hard Prompts | 1435 | 1220 |
| Chess Puzzles | — | 1% |
| LiveBench Reasoning | — | 47.8% |
| LiveBench Data Analysis | — | 45.2% |
| Epoch Capabilities Index | — | 130.42 |
| LiveBench | — | 41.6% |
Math Grok 4.1 leads
Grok 4.1: 38.9 (#120), Phi-4: 20.8 (#285)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Math | 1422 | 1246 |
| OTIS Mock AIME 2024-2025 | — | 13.8% |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
Knowledge Grok 4.1 leads
Grok 4.1: 39.5 (#133), Phi-4: 32.6 (#209)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Expert | 1417 | 1203 |
| GPQA Diamond | — | 56.1% |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| MMLU | — | 84.8% |
Multilingual Grok 4.1 leads
Grok 4.1: 53.4 (#68), Phi-4: 37.2 (#237)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Non-English | 1425 | 1197 |
| LMArena Chinese | 1465 | 1212 |
| LMArena French | 1448 | 1224 |
| LMArena German | 1446 | 1222 |
| LMArena Japanese | 1397 | 1158 |
| LMArena Korean | 1407 | 1151 |
| LMArena Russian | 1434 | 1209 |
| LMArena Spanish | 1438 | 1234 |
Instruction Following Grok 4.1 leads
Grok 4.1: 73.8 (#111), Phi-4: 60.4 (#251)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1400 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
Long Context Grok 4.1 leads
Grok 4.1: 43.2 (#100), Phi-4: 36.9 (#226)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1416 | 1217 |
Writing & Preference Grok 4.1 leads
Grok 4.1: 62.4 (#75), Phi-4: 40.5 (#244)
| Benchmark | Grok 4.1 | Phi-4 |
|---|---|---|
| LMArena Text | 1437 | 1217 |
| LMArena Creative Writing | 1411 | 1182 |
| LMArena Multi-Turn | 1437 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Grok 4.1 better than Phi-4?
Grok 4.1 is the stronger model overall, scoring 41.5 to 31.2 on the Noometry Index.
Is Grok 4.1 or Phi-4 better for coding?
They score almost the same on coding (33.7 vs 34.4); test both on your own repository before choosing.
How many benchmarks do Grok 4.1 and Phi-4 share?
17 benchmarks have published results for both models. Grok 4.1 has 19 scored results on Noometry and Phi-4 has 37.