Model comparison
Magistral Medium vs Phi-4
Magistral Medium is the stronger model overall, scoring 35.2 to 31.2 on the Noometry Index. Phi-4 costs 31× less per token, which makes it the better buy when Magistral Medium's lead doesn't matter for your workload.
Last verified . 17 shared benchmarks.
Mistral AI
MicrosoftSummary
- They share 17 benchmarks with published results for both. Magistral Medium scores higher in 7 categories and Phi-4 in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Magistral Medium leads 35.1 to 20.8.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $2 / $5 for Magistral Medium.
- Magistral Medium accepts more context: 262K tokens versus 128K.
Side by side
| Magistral Medium | Phi-4 | |
|---|---|---|
| Provider | Mistral AI | Microsoft |
| Noometry Index | 35.2 | 31.2 |
| Released | 2025-03-17 | 2024-12-11 |
| Weights | Open | Open |
| Context window | 262K | 128K |
| Max output | 16K | 4K |
| Input $ / M tokens | $2 | $0.07 |
| Output $ / M tokens | $5 | $0.14 |
| Results tracked | 22 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Magistral Medium leads
Magistral Medium: 39.1 (#161), Phi-4: 34.4 (#239)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Coding | 1319 | 1231 |
| SciCode | 39.2% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
Agentic & Tool Use Not comparable
Magistral Medium: —, Phi-4: 22.8 (#128)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.8% |
| BALROG | — | 11.6% |
Reasoning Phi-4 leads
Magistral Medium: 8.6 (#348), Phi-4: 17.7 (#291)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Hard Prompts | 1267 | 1220 |
| ARC-AGI-2 | 0% | — |
| Kagi LLM Benchmark | 16.2% | — |
| ARC-AGI-1 | 6.1% | — |
| CritPt | 0.3% | — |
| Chess Puzzles | — | 1% |
| LiveBench Reasoning | — | 47.8% |
| LiveBench Data Analysis | — | 45.2% |
| Epoch Capabilities Index | — | 130.42 |
| LiveBench | — | 41.6% |
Math Magistral Medium leads
Magistral Medium: 35.1 (#189), Phi-4: 20.8 (#285)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Math | 1250 | 1246 |
| OTIS Mock AIME 2024-2025 | — | 13.8% |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
Knowledge Too close to call
Magistral Medium: 33.5 (#202), Phi-4: 32.6 (#209)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Expert | 1223 | 1203 |
| GPQA Diamond | — | 56.1% |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| MMLU | — | 84.8% |
Multilingual Magistral Medium leads
Magistral Medium: 39.6 (#224), Phi-4: 37.2 (#237)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Non-English | 1232 | 1197 |
| LMArena Chinese | 1227 | 1212 |
| LMArena French | 1267 | 1224 |
| LMArena German | 1248 | 1222 |
| LMArena Japanese | 1175 | 1158 |
| LMArena Korean | 1125 | 1151 |
| LMArena Russian | 1224 | 1209 |
| LMArena Spanish | 1271 | 1234 |
Instruction Following Magistral Medium leads
Magistral Medium: 66.0 (#211), Phi-4: 60.4 (#251)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1254 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
Long Context Magistral Medium leads
Magistral Medium: 39.3 (#183), Phi-4: 36.9 (#226)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1295 | 1217 |
Writing & Preference Magistral Medium leads
Magistral Medium: 46.3 (#219), Phi-4: 40.5 (#244)
| Benchmark | Magistral Medium | Phi-4 |
|---|---|---|
| LMArena Text | 1255 | 1217 |
| LMArena Creative Writing | 1245 | 1182 |
| LMArena Multi-Turn | 1275 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Magistral Medium better than Phi-4?
Magistral Medium is the stronger model overall, scoring 35.2 to 31.2 on the Noometry Index. Phi-4 costs 31× less per token, which makes it the better buy when Magistral Medium's lead doesn't matter for your workload.
Which is cheaper, Magistral Medium or Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Magistral Medium lists at $2 and $5.
Is Magistral Medium or Phi-4 better for coding?
Magistral Medium scores higher on coding benchmarks: 39.1 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
Magistral Medium does, with 262K tokens against 128K.
How many benchmarks do Magistral Medium and Phi-4 share?
17 benchmarks have published results for both models. Magistral Medium has 22 scored results on Noometry and Phi-4 has 37.