Model comparison
Grok 3 vs Phi-4
Grok 3 is the stronger model overall, scoring 39.9 to 31.2 on the Noometry Index.
Last verified . 25 shared benchmarks.
MicrosoftSummary
- They share 25 benchmarks with published results for both. Grok 3 scores higher in 8 categories and Phi-4 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 3 leads 38.0 to 20.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 55.6% for Grok 3 and 13.8% for Phi-4.
- Phi-4 has downloadable open weights; the other is API-only.
Side by side
| Grok 3 | Phi-4 | |
|---|---|---|
| Provider | xAI | Microsoft |
| Noometry Index | 39.9 | 31.2 |
| Released | 2025-04-09 | 2024-12-11 |
| Weights | Proprietary | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.07 |
| Output $ / M tokens | — | $0.14 |
| Results tracked | 40 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 3 leads
Grok 3: 41.9 (#115), Phi-4: 34.4 (#239)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| LMArena Coding | 1432 | 1231 |
| Aider Polyglot | 53.3% | — |
| WeirdML | 37.2% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
Agentic & Tool Use Grok 3 leads
Grok 3: 30.5 (#76), Phi-4: 22.8 (#128)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| BALROG | 29.5% | 11.6% |
| Berkeley Function Calling Leaderboard | — | 28.8% |
Reasoning Phi-4 leads
Grok 3: 13.7 (#333), Phi-4: 17.7 (#291)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| LMArena Hard Prompts | 1434 | 1220 |
| Epoch Capabilities Index | 138.33 | 130.42 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 36.1% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 5.5% | — |
| Chess Puzzles | — | 1% |
| LiveBench Reasoning | — | 47.8% |
| LiveBench Data Analysis | — | 45.2% |
| LiveBench | — | 41.6% |
Math Grok 3 leads
Grok 3: 38.0 (#145), Phi-4: 20.8 (#285)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 55.6% | 13.8% |
| LMArena Math | 1391 | 1246 |
| MATH Level 5 | 88.7% | 64.9% |
| Omni-MATH | 46.4% | — |
| LiveBench Math | — | 42% |
| FrontierMath (Feb 2025 set) | 3.8% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Grok 3 leads
Grok 3: 46.2 (#82), Phi-4: 32.6 (#209)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| GPQA Diamond | 75.8% | 56.1% |
| Confabulations | 14.2% | 29.4% |
| Vectara Hallucination Rate | 5.8% | 3.7% |
| LMArena Expert | 1421 | 1203 |
| MMLU-Pro | 78.8% | — |
| GPQA (HELM) | 65% | — |
| MMLU | — | 84.8% |
Multilingual Grok 3 leads
Grok 3: 52.3 (#87), Phi-4: 37.2 (#237)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| LMArena Non-English | 1410 | 1197 |
| LMArena Chinese | 1448 | 1212 |
| LMArena French | 1460 | 1224 |
| LMArena German | 1431 | 1222 |
| LMArena Japanese | 1387 | 1158 |
| LMArena Korean | 1373 | 1151 |
| LMArena Russian | 1416 | 1209 |
| LMArena Spanish | 1417 | 1234 |
Instruction Following Grok 3 leads
Grok 3: 75.0 (#73), Phi-4: 60.4 (#251)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1409 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
| IFEval | 88.4% | — |
Long Context Grok 3 leads
Grok 3: 38.7 (#192), Phi-4: 36.9 (#226)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1439 | 1217 |
| Fiction.LiveBench | 58.3% | — |
Writing & Preference Grok 3 leads
Grok 3: 55.8 (#141), Phi-4: 40.5 (#244)
| Benchmark | Grok 3 | Phi-4 |
|---|---|---|
| LMArena Text | 1426 | 1217 |
| LMArena Creative Writing | 1414 | 1182 |
| Short-Story Creative Writing | 76.4% | 62.6% |
| LMArena Multi-Turn | 1425 | 1206 |
| EQ-Bench Creative Writing | 1186 | — |
| WildBench | 84.9% | — |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Grok 3 better than Phi-4?
Grok 3 is the stronger model overall, scoring 39.9 to 31.2 on the Noometry Index.
Is Grok 3 or Phi-4 better for coding?
Grok 3 scores higher on coding benchmarks: 41.9 versus 34.4 in the Noometry coding category.
How many benchmarks do Grok 3 and Phi-4 share?
25 benchmarks have published results for both models. Grok 3 has 40 scored results on Noometry and Phi-4 has 37.