Model comparison

Phi 3 Medium 4k Instruct vs Phi 3 Mini 4k Instruct

Phi 3 Medium 4k Instruct is the stronger model overall, scoring 33.0 to 27.9 on the Noometry Index.

Last verified . 17 shared benchmarks.

Phi 3 Medium 4k Instruct Microsoft

33.0

Rank #250 Confirmed

Phi 3 Mini 4k Instruct Microsoft

27.9

Rank #328 Confirmed

Summary

  • They share 17 benchmarks with published results for both. Phi 3 Medium 4k Instruct scores higher in 8 categories and Phi 3 Mini 4k Instruct in 0 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in instruction following, where Phi 3 Medium 4k Instruct leads 57.6 to 47.7.

Side by side

Phi 3 Medium 4k Instruct and Phi 3 Mini 4k Instruct specifications
Phi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
ProviderMicrosoftMicrosoft
Noometry Index33.027.9
Released—2024-04-23
WeightsOpenOpen
Context window——
Max output——
Input $ / M tokens——
Output $ / M tokens——
Results tracked1735

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 32.9 (#268), Phi 3 Mini 4k Instruct: 26.6 (#323)

Coding benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Coding11301093
LiveBench Coding—15.5%
HumanEval+—59.1%
MBPP+—54.2%

Reasoning Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 21.7 (#214), Phi 3 Mini 4k Instruct: 14.1 (#328)

Reasoning benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Hard Prompts11271072
Chess Puzzles—0%
LiveBench Reasoning—26.8%
LiveBench Data Analysis—34.7%
Adversarial NLI—52.8%
BIG-Bench Hard—71.7%
HellaSwag—76.7%
LiveBench—22.4%
WinoGrande—70.8%

Math Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 33.4 (#202), Phi 3 Mini 4k Instruct: 26.6 (#257)

Math benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Math11731111
LiveBench Math—15.7%

Knowledge Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 30.2 (#229), Phi 3 Mini 4k Instruct: 28.5 (#246)

Knowledge benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Expert11081045
ARC (AI2) Challenge—84.9%
MMLU—68.8%
OpenBookQA—88%
TriviaQA—64%

Multilingual Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 30.6 (#263), Phi 3 Mini 4k Instruct: 26.3 (#280)

Multilingual benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Non-English10941021
LMArena Chinese11081021
LMArena French11101076
LMArena German11011044
LMArena Japanese1042935
LMArena Korean954905
LMArena Russian11451022
LMArena Spanish10951085

Instruction Following Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 57.6 (#268), Phi 3 Mini 4k Instruct: 47.7 (#303)

Instruction Following benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Instruction Following11131053
LiveBench Instruction Following—39.1%

Long Context Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 34.0 (#255), Phi 3 Mini 4k Instruct: 31.7 (#276)

Long Context benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Longer Query11201044

Writing & Preference Phi 3 Medium 4k Instruct leads

Phi 3 Medium 4k Instruct: 34.1 (#272), Phi 3 Mini 4k Instruct: 27.6 (#300)

Writing & Preference benchmarks
BenchmarkPhi 3 Medium 4k InstructPhi 3 Mini 4k Instruct
LMArena Text11381073
LMArena Creative Writing11071037
LMArena Multi-Turn10881018
LiveBench Language—9.2%

Frequently asked questions

Is Phi 3 Medium 4k Instruct better than Phi 3 Mini 4k Instruct?

Phi 3 Medium 4k Instruct is the stronger model overall, scoring 33.0 to 27.9 on the Noometry Index.

Is Phi 3 Medium 4k Instruct or Phi 3 Mini 4k Instruct better for coding?

Phi 3 Medium 4k Instruct scores higher on coding benchmarks: 32.9 versus 26.6 in the Noometry coding category.

How many benchmarks do Phi 3 Medium 4k Instruct and Phi 3 Mini 4k Instruct share?

17 benchmarks have published results for both models. Phi 3 Medium 4k Instruct has 17 scored results on Noometry and Phi 3 Mini 4k Instruct has 35.

Related comparisons

Go deeper