Model profile
Microsoft Azure

Microsoft: Phi 4

Microsoft: Phi 4 is a budget-priced text-first model from Microsoft Azure with balanced runtime profile, compact context posture, and the clearest fit around long-context research / coding.

Best for: Long-context research / CodingBalanced latencyCompact contextBudget pricing
Intelligence
10.4

Benchmark blend

Coding
11.2

Dev workflow signal

Context
16K Tokens

Compact

Input Price
$0.13

Budget tier

Decision snapshot
35

Microsoft: Phi 4 currently reads as a budget text-first option with compact context and a balanced runtime profile.

Overall profile
Use-case specific
Best for
Long-context research / Coding
Latency tier
Balanced
Price tier
Budget
Source coverage
OpenRouterArtificial Analysis

Decision Strip

Decision rail before the raw tables

Core buy-side signals stay in one pass. The rest of the page expands only after intelligence, speed, context, and price are clear.

Intelligence
10.4
10

General reasoning and benchmark headroom.

Limited
Speed
37 tok/s
61

TTFT 0.49s

Competitive
Context
16K Tokens
28

How much prompt and task state can stay in view.

Limited
Price
$0.13
86

$0.50 output / 1M

Efficient

Editorial Profile

Microsoft: Phi 4 in one narrative

Positioning, tradeoffs, and fit are consolidated into one read instead of repeating the same story across separate cards.

Use-case specificCoding score 11Math score 18

[Microsoft Research](/microsoft) Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. At 14 billion...

Identity

Microsoft Azure text-first profile

Positioning

Long-context research / Coding with compact context and balanced runtime.

Cost posture

Efficient spend profile. More comfortable for sustained prompt volume if the capability fit is right.

Strengths
  • The available source data suggests a balanced profile rather than one dominant edge.

Tradeoffs
  • Budget-friendly input pricing is a strength, but raw capability may vary by workload.

  • Latency is balanced rather than ultra-fast, which is fine for most workflows but not the snappiest tier.

  • Current metadata points to a text-first profile rather than a broad multimodal one.

  • Context window is more comfortable for focused tasks than extremely long sessions.

Best fit
  • Focused chat, retrieval-augmented flows, and narrower production tasks.

Explore Next

Similar profiles worth opening next

Microsoft Azure

Phi-3 Mini Instruct 3.8B

Intelligence
10.1
Context
N/A
Input Price
$0.00
Microsoft Azure

Phi-4 Multimodal Instruct

Intelligence
10.0
Context
N/A
Input Price
$0.00
Microsoft Azure

Phi-4 Mini Instruct

Intelligence
8.4
Context
N/A
Input Price
$0.00

Benchmarks

Grouped by job-to-be-done

Only benchmark categories with actual signal are shown. Secondary values stay as simple definitions instead of nested micro-cards.

General intelligence

Broad reasoning, knowledge depth, and flagship benchmark posture.

Intelligence Index
10.4
MMLU Pro
71.4%
GPQA
57.5%
HLE
4.1%
Coding

Software implementation, debugging quality, and coding benchmark signal.

Coding Index
11.2
LiveCodeBench
0.231
SciCode
26.0%
Math

Formal reasoning, structured problem solving, and competition-style math.

Math Index
18.0
AIME
14.3%
AIME 2025
18.0%
Math 500
81.0%
Agent / tool use

Long-horizon execution quality and interactive benchmark evidence.

IFBench
23.5%
TAU2
0.0%
TerminalBench Hard
3.8%
LCR
0.0%

Specs & Pricing

Technical snapshot and cost posture

Specs stay neutral, pricing gets emphasis through values rather than extra containers. Raw provider internals remain in metadata at the end.

Technical snapshot
Context Window
16K Tokens
Vision
Text-first
Modalities
text
Tokenizer
Other
Max Completion
16384
Moderation
No
Supported Parameters
frequency_penaltylogprobsmax_tokensmin_ppresence_penaltyrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetop_ktop_logprobstop_p
Input Modalities
text
Output Modalities
text
Price architecture
Input
per 1M input tokens
$0.13
Output
per 1M output tokens
$0.50
Blended
AA 3:1 mix
$0.22

This model is relatively efficient on price. It is the easier fit when sustained prompt volume matters.

Metadata

Raw source tables at the end

Verification details remain available, but the page no longer forces them ahead of the editorial read.