Phi-4 Mini
Microsoft · 1 configuration
Intelligence6.3
Input / 1M$0
Output / 1M$0
Context128K
Tariff: Microsoft · Source: artificial-analysis · USD per million tokens
Thinking configuration
What does more thinking buy you?
Compare this model’s measured thinking configurations.
Scroll the chart ↔ · tap a point for details
Efficient frontier: best measured score at each cost1 measured configurations · 0 without both values
Selected: thinking default · Click a point to switch configuration
Compact GPT model for low-latency assistance and high-volume workloads
Intelligence IndexArtificial Analysis composite index, not a percentage. Different versions of the index are not directly comparable.
6.3 pointsGPQA DiamondGraduate-level scientific questions. Measures specialist reasoning rather than web-search quality.
33.1 %Humanity’s Last ExamDifficult questions across many disciplines, compared under the same evaluation setup.
4.5 %Terminal-Bench 2.1Agent tasks in a terminal environment. Results also depend on the evaluation harness.
0.4 %τ²-BenchTool use and interaction in assistance scenarios. Does not replace an evaluation of your integration.
8.2 %IFBenchVerifiable instruction following. A proxy for precision rather than a benchmark of creative style.
21.1 %Omniscience accuracyFactual accuracy in Artificial Analysis's Omniscience evaluation.
9.5 %Specifications from the catalog
Inputtext
Outputtext
Tool callingNo
Structured outputNo
Speed44.8 tok/s
First token0.82 s
Maximum output16.4K
Catalog listing date2025-07-26