modelbenchmark.io

Phi-4

Microsoft · phi-4

Compare

Open-weight instruction model for adaptable chat and self-hosted production workloads

Specification

most-agreed values

Context
128K
Max output
4K
Released
2024-12-11
Knowledge cutoff
2023-10
Retires
2027-01-17
Open weights
yes
Input
text
Output
text

Price

US dollars per million tokens · most-agreed

Input
$0.125
Output
$0.50

Quality

3 benchmarks · 1 source

Composite
39th
percentile of 344 scored models
Rank
211 / 344
±0.06 sd
Evidence
3 × 1
one source, or fewer than 3 benchmarks
Effort range
—
one configuration only
reasoning41st1/4 bench
math37th2/7 bench

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
MATH level 564.9 ±1.1defaultEpoch AI2025-01-31
GPQA diamond56.1 ±2.6defaultEpoch AI2025-01-31
OTIS Mock AIME 2024-202513.8 ±3.7defaultEpoch AI2025-02-25

The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.

Available from 6 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
kilo$0.056$0.112——16K15K—
deepinfra$0.07$0.14——16K16K—
openrouter$0.07$0.14——16K15K—
azure$0.125$0.50——128K4K—
azure_ai$0.125$0.50——16K16K2027-01-17
azure-cognitive-services$0.125$0.50——128K4K—

Price history

append-only observations · a listing writes a row only when its price moves

DateHostField$/M
2026-09-02kiloInput$0.07
2026-10-16kiloInput$0.056
2026-09-02openrouterInput$0.07
2026-09-19openrouterInput$0.07
2026-10-10openrouterInput$0.07
2026-09-02kiloOutput$0.14
2026-10-16kiloOutput$0.112
2026-09-02openrouterOutput$0.14
2026-09-19openrouterOutput$0.14
2026-10-10openrouterOutput$0.14

More from Microsoft

most-hosted first