Rankings / Microsoft Research
Phi 4
released 2024-12-12
BenchAtlas Index
as of 2026-09-11
base configuration
rank #445
12 families · 5 categories · high
Benchmark evidence
42 results
agentic coding
| τ²-Bench | 0.0 % | independent | Artificial Analysis |
coding
| LiveCodeBench v6 implementation=artificial-analysis | 23.1 % | independent | Artificial Analysis | |
| SciCode | 26.0 % | independent | Artificial Analysis |
external indices
| Intelligence Index v4.1 | 5.9 points | independent | Artificial Analysis | |
| AA Math Index | 18.0 points | independent | Artificial Analysis | |
| ECI | 130.4 points | independent | Epoch AI Benchmarking Hub |
human preference
| Coding (style control)older version arena=text · category=coding · style_control=true | 1306.2 1296.3–1316.1 | community | LMArena Leaderboard Dataset | |
| Industry Legal And Government (style control)older version arena=text · category=industry_legal_and_government · style_control=true | 1298.8 1282.1–1315.4 | community | LMArena Leaderboard Dataset | |
| Industry Software And It Services (style control)older version arena=text · category=industry_software_and_it_services · style_control=true | 1289.8 1281.7–1297.9 | community | LMArena Leaderboard Dataset | |
| Hard Prompts English (style control)older version arena=text · category=hard_prompts_english · style_control=true | 1289.1 1279.7–1298.5 | community | LMArena Leaderboard Dataset | |
| Industry Mathematical (style control)older version arena=text · category=industry_mathematical · style_control=true | 1278.3 1266.7–1289.9 | community | LMArena Leaderboard Dataset | |
| Hard Prompts (style control)older version arena=text · category=hard_prompts · style_control=true | 1277.5 1269.7–1285.3 | community | LMArena Leaderboard Dataset | |
| Industry Medicine And Healthcare (style control)older version arena=text · category=industry_medicine_and_healthcare · style_control=true | 1272.6 1253.2–1292.0 | community | LMArena Leaderboard Dataset | |
| English (style control)older version arena=text · category=english · style_control=true | 1271.8 1266.0–1277.6 | community | LMArena Leaderboard Dataset | |
| Expert (style control)older version arena=text · category=expert · style_control=true | 1267.9 1250.6–1285.2 | community | LMArena Leaderboard Dataset | |
| Chinese (style control)older version arena=text · category=chinese · style_control=true | 1267.2 1251.6–1282.8 | community | LMArena Leaderboard Dataset | |
| Longer Query (style control)older version arena=text · category=longer_query · style_control=true | 1266.6 1255.7–1277.4 | community | LMArena Leaderboard Dataset | |
| Math (style control)older version arena=text · category=math · style_control=true | 1264.7 1254.2–1275.2 | community | LMArena Leaderboard Dataset | |
| Industry Life And Physical And Social Science (style control)older version arena=text · category=industry_life_and_physical_and_social_science · style_control=true | 1264.6 1254.9–1274.3 | community | LMArena Leaderboard Dataset | |
| Overall (style control) arena=text · category=overall · style_control=true | 1256.0 1251.4–1260.6 | community | LMArena Leaderboard Dataset | |
| Industry Business And Management And Financial Operations (style control)older version arena=text · category=industry_business_and_management_and_financial_operations · style_control=true | 1255.9 1243.8–1268.0 | community | LMArena Leaderboard Dataset | |
| German (style control)older version arena=text · category=german · style_control=true | 1255.7 1229.9–1281.5 | community | LMArena Leaderboard Dataset | |
| Russian (style control)older version arena=text · category=russian · style_control=true | 1245.7 1234.4–1257.0 | community | LMArena Leaderboard Dataset | |
| Instruction Following (style control)older version arena=text · category=instruction_following · style_control=true | 1245.2 1238.6–1251.9 | community | LMArena Leaderboard Dataset | |
| Multi Turn (style control)older version arena=text · category=multi_turn · style_control=true | 1241.9 1231.7–1252.2 | community | LMArena Leaderboard Dataset | |
| Non English (style control)older version arena=text · category=non_english · style_control=true | 1232.7 1226.2–1239.1 | community | LMArena Leaderboard Dataset | |
| Industry Writing And Literature And Language (style control)older version arena=text · category=industry_writing_and_literature_and_language · style_control=true | 1228.8 1221.0–1236.6 | community | LMArena Leaderboard Dataset | |
| Creative Writing (style control)older version arena=text · category=creative_writing · style_control=true | 1209.8 1200.1–1219.5 | community | LMArena Leaderboard Dataset | |
| Industry Entertainment And Sports And Media (style control)older version arena=text · category=industry_entertainment_and_sports_and_media · style_control=true | 1202.0 1192.9–1211.1 | community | LMArena Leaderboard Dataset | |
| Japanese (style control)older version arena=text · category=japanese · style_control=true | 1194.1 1167.3–1220.9 | community | LMArena Leaderboard Dataset | |
| Exclude Ties (style control)older version arena=text · category=exclude_ties · style_control=true | 1156.5 1148.7–1164.3 | community | LMArena Leaderboard Dataset |
knowledge science
| GPQA Diamond implementation=artificial-analysis | 57.5 % | independent | Artificial Analysis | |
| GPQA Diamond | 56.1 % | independent | Epoch AI Benchmarking Hub 2025-01-31 | |
| Humanity's Last Exam implementation=artificial-analysis | 3.8 % | independent | Artificial Analysis | |
| MMLU-Pro implementation=artificial-analysis | 71.4 % | independent | Artificial Analysis |
long context instruction
| AA-LCR | 0.0 % | independent | Artificial Analysis | |
| IFBench | 23.5 % | independent | Artificial Analysis |
reasoning math
| AIME year=2025 · implementation=artificial-analysis | 18.0 % | independent | Artificial Analysis | |
| AIME implementation=artificial-analysis | 14.3 % | independent | Artificial Analysis | |
| MATH-500 implementation=artificial-analysis | 81.0 % | independent | Artificial Analysis | |
| MATH Level 5 | 64.9 % | independent | Epoch AI Benchmarking Hub 2025-01-31 | |
| OTIS Mock AIME 2024–2025 | 13.8 % | independent | Epoch AI Benchmarking Hub 2025-02-25 |
Agent + model results
systems, not bare-model scores
| agent + model Artificial Analysis harness + Phi 4 | Terminal-Bench Hard | 3.8 % | independent | Artificial Analysis |
These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.