Model comparison
Grok 4 vs phi-3-medium 14B
Grok 4 is the stronger model overall, scoring 48.1 to 29.7 on the Noometry Index.
Last verified . 2 shared benchmarks.
MicrosoftSummary
- They share 2 benchmarks with published results for both. Grok 4 scores higher in 3 categories and phi-3-medium 14B in 0 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok 4 leads 53.8 to 9.1.
- The biggest single-benchmark swing is GPQA Diamond: 87% for Grok 4 and 27.6% for phi-3-medium 14B.
- phi-3-medium 14B has downloadable open weights; the other is API-only.
Side by side
| Grok 4 | phi-3-medium 14B | |
|---|---|---|
| Provider | xAI | Microsoft |
| Noometry Index | 48.1 | 29.7 |
| Released | 2025-07-09 | 2024-04-23 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 48 | 13 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4 leads
Grok 4: 50.3 (#46), phi-3-medium 14B: 36.8 (#201)
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| Aider Polyglot | 79.6% | — |
| WeirdML | 45.7% | — |
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1408 | — |
| BigCodeBench Complete | — | 48.7% |
Agentic & Tool Use Not comparable
Grok 4: 32.3 (#68), phi-3-medium 14B: —
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| Terminal-Bench | 27.2% | — |
| Berkeley Function Calling Leaderboard | 63% | — |
| GDPval | 21.1% | — |
| Cybench | 43% | — |
| DeepResearch Bench | 47.3% | — |
| BALROG | 43.6% | — |
| LMArena Search | 1142 | — |
| METR Time Horizons | 66.6% | — |
Reasoning Not comparable
Grok 4: 36.7 (#65), phi-3-medium 14B: —
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| Epoch Capabilities Index | 146.44 | 121.23 |
| ARC-AGI-2 | 16% | — |
| SimpleBench | 60.5% | — |
| Kagi LLM Benchmark | 73.6% | — |
| ARC-AGI-1 | 66.7% | — |
| Chess Puzzles | 28% | — |
| LMArena Hard Prompts | 1409 | — |
| Adversarial NLI | — | 55.8% |
| BIG-Bench Hard | — | 81.4% |
| ForecastBench | 60.9 | — |
| HellaSwag | — | 82.4% |
| WinoGrande | — | 81.5% |
Math Grok 4 leads
Grok 4: 48.4 (#64), phi-3-medium 14B: 27.3 (#250)
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 84% | — |
| Omni-MATH | 60.3% | — |
| LMArena Math | 1422 | — |
| MATH Level 5 | — | 17.6% |
| FrontierMath (Feb 2025 set) | 19.7% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Grok 4 leads
Grok 4: 53.8 (#55), phi-3-medium 14B: 9.1 (#306)
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| GPQA Diamond | 87% | 27.6% |
| MMLU-Pro | 85.1% | — |
| Confabulations | 12.4% | — |
| GPQA (HELM) | 72.7% | — |
| LMArena Expert | 1415 | — |
| ARC (AI2) Challenge | — | 91.6% |
| MMLU | — | 78% |
| OpenBookQA | — | 87.4% |
| TriviaQA | — | 73.9% |
Multimodal Not comparable
Grok 4: 33.7 (#94), phi-3-medium 14B: —
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| LMArena Vision | 1210 | — |
| GeoBench | 45% | — |
Multilingual Not comparable
Grok 4: 51.8 (#103), phi-3-medium 14B: —
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| LMArena Non-English | 1403 | — |
| LMArena Chinese | 1427 | — |
| LMArena French | 1418 | — |
| LMArena German | 1429 | — |
| LMArena Japanese | 1394 | — |
| LMArena Korean | 1377 | — |
| LMArena Russian | 1410 | — |
| LMArena Spanish | 1420 | — |
Instruction Following Not comparable
Grok 4: 79.2 (#5), phi-3-medium 14B: —
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| IFEval | 94.9% | — |
| LMArena Instruction Following | 1387 | — |
Long Context Not comparable
Grok 4: 63.1 (#4), phi-3-medium 14B: —
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| Fiction.LiveBench | 94.4% | — |
| LMArena Longer Query | 1409 | — |
Writing & Preference Not comparable
Grok 4: 58.5 (#116), phi-3-medium 14B: —
| Benchmark | Grok 4 | phi-3-medium 14B |
|---|---|---|
| LMArena Text | 1411 | — |
| LMArena Creative Writing | 1397 | — |
| Short-Story Creative Writing | 76.9% | — |
| WildBench | 79.7% | — |
| LMArena Multi-Turn | 1416 | — |
Frequently asked questions
Is Grok 4 better than phi-3-medium 14B?
Grok 4 is the stronger model overall, scoring 48.1 to 29.7 on the Noometry Index.
Is Grok 4 or phi-3-medium 14B better for coding?
Grok 4 scores higher on coding benchmarks: 50.3 versus 36.8 in the Noometry coding category.
How many benchmarks do Grok 4 and phi-3-medium 14B share?
2 benchmarks have published results for both models. Grok 4 has 48 scored results on Noometry and phi-3-medium 14B has 13.