Model comparison
Llama 4 Scout vs Phi-4
Phi-4 is the stronger model overall, scoring 31.2 to 27.7 on the Noometry Index.
Last verified . 24 shared benchmarks.
Meta
MicrosoftSummary
- They share 24 benchmarks with published results for both. Llama 4 Scout scores higher in 3 categories and Phi-4 in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where Phi-4 leads 34.4 to 20.2.
- The biggest single-benchmark swing is BigCodeBench Complete: 43.1% for Llama 4 Scout and 55.4% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.10 / $0.30 for Llama 4 Scout.
Side by side
| Llama 4 Scout | Phi-4 | |
|---|---|---|
| Provider | Meta | Microsoft |
| Noometry Index | 27.7 | 31.2 |
| Released | 2025-04-05 | 2024-12-11 |
| Weights | Open | Open |
| Context window | 128K | 128K |
| Max output | 4K | 4K |
| Input $ / M tokens | $0.10 | $0.07 |
| Output $ / M tokens | $0.30 | $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
Llama 4 Scout: 20.2 (#339), Phi-4: 34.4 (#239)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| LMArena Coding | 1286 | 1231 |
| BigCodeBench Complete | 43.1% | 55.4% |
| SWE-bench Verified (bash only) | 9.1% | — |
| SciCode | 17% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
Agentic & Tool Use Llama 4 Scout leads
Llama 4 Scout: 24.6 (#119), Phi-4: 22.8 (#128)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 28.1% | 28.8% |
| BALROG | — | 11.6% |
Reasoning Phi-4 leads
Llama 4 Scout: 9.1 (#345), Phi-4: 17.7 (#291)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| LMArena Hard Prompts | 1266 | 1220 |
| Epoch Capabilities Index | 129.64 | 130.42 |
| ARC-AGI-2 | 0% | — |
| Kagi LLM Benchmark | 36.9% | — |
| ARC-AGI-1 | 0.5% | — |
| CritPt | 0% | — |
| Chess Puzzles | — | 1% |
| LiveBench Reasoning | — | 47.8% |
| DTBench | 57.9% | — |
| LiveBench Data Analysis | — | 45.2% |
| LMCA | 12% | — |
| ForecastBench | 57.5 | — |
| LiveBench | — | 41.6% |
Math Phi-4 leads
Llama 4 Scout: 19.6 (#286), Phi-4: 20.8 (#285)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 7.8% | 13.8% |
| LMArena Math | 1287 | 1246 |
| MATH Level 5 | 62.3% | 64.9% |
| Omni-MATH | 37.3% | — |
| LiveBench Math | — | 42% |
| FrontierMath (Feb 2025 set) | 0% | — |
Knowledge Too close to call
Llama 4 Scout: 31.9 (#217), Phi-4: 32.6 (#209)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| GPQA Diamond | 51.8% | 56.1% |
| Vectara Hallucination Rate | 7.7% | 3.7% |
| LMArena Expert | 1235 | 1203 |
| MMLU-Pro | 74.2% | — |
| Confabulations | — | 29.4% |
| GPQA (HELM) | 50.7% | — |
| MMLU | — | 84.8% |
Multimodal Not comparable
Llama 4 Scout: 32.2 (#102), Phi-4: —
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| LMArena Vision | 1118 | — |
| SpatialViz-Bench | 34.2% | — |
Multilingual Llama 4 Scout leads
Llama 4 Scout: 41.0 (#212), Phi-4: 37.2 (#237)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| LMArena Non-English | 1252 | 1197 |
| LMArena Chinese | 1255 | 1212 |
| LMArena French | 1282 | 1224 |
| LMArena German | 1272 | 1222 |
| LMArena Japanese | 1206 | 1158 |
| LMArena Korean | 1207 | 1151 |
| LMArena Russian | 1263 | 1209 |
| LMArena Spanish | 1278 | 1234 |
Instruction Following Llama 4 Scout leads
Llama 4 Scout: 65.8 (#217), Phi-4: 60.4 (#251)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1248 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
| IFEval | 81.8% | — |
Long Context Phi-4 leads
Llama 4 Scout: 27.5 (#294), Phi-4: 36.9 (#226)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1265 | 1217 |
| Fiction.LiveBench | 36% | — |
Writing & Preference Phi-4 leads
Llama 4 Scout: 37.0 (#261), Phi-4: 40.5 (#244)
| Benchmark | Llama 4 Scout | Phi-4 |
|---|---|---|
| LMArena Text | 1279 | 1217 |
| LMArena Creative Writing | 1249 | 1182 |
| LMArena Multi-Turn | 1280 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| EQ-Bench Creative Writing | 783 | — |
| WildBench | 78% | — |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Llama 4 Scout better than Phi-4?
Phi-4 is the stronger model overall, scoring 31.2 to 27.7 on the Noometry Index.
Which is cheaper, Llama 4 Scout or Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Llama 4 Scout lists at $0.10 and $0.30.
Is Llama 4 Scout or Phi-4 better for coding?
Phi-4 scores higher on coding benchmarks: 34.4 versus 20.2 in the Noometry coding category.
Which has the bigger context window?
Both accept 128K tokens.
How many benchmarks do Llama 4 Scout and Phi-4 share?
24 benchmarks have published results for both models. Llama 4 Scout has 43 scored results on Noometry and Phi-4 has 37.