Microsoft
Efficient

Phi-3 Medium

Balanced

14B Phi-3, Microsoft's mid-size SLM. Single-GPU friendly with 128K context.

Phi-3 Medium is a efficient AI model from Microsoft. It costs $0.1700 per million input tokens and $0.6800 per million output tokens (blended $0.3230/M), with a 128,000-token context window.

RatesUSD per million tokens. Blend is 70% input, 30% output, our calculation.

INPUT
$0.1700/M
per million input tokens
OUTPUT
$0.6800/M
per million output tokens
BLENDED 70/30
$0.3230/M
unchanged since 3 May
CONTEXT
128,000
tokens

Hand-maintained from Microsoft's published pricing; no check date on record.

What it is good at
  • Single-GPU
  • Open weights (MIT)
  • 128K context
Typical use cases
  • On-prem reasoning
  • Edge inference research

Benchmarks

vs. best public score
MMLU78%
Multitask academic knowledge across 57 subjects. Source: Microsoft (self-reported), MMLU, read 2026-11-10.
MATH53.1%
High-school competition math problems. Source: Microsoft (self-reported), MATH, read 2026-11-10.
HumanEval62.2%
Python function synthesis from docstrings. Source: Microsoft (self-reported), HumanEval (0-shot, phi-3-medium-instruct), read 2026-11-10.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

More from Microsoft

See all 8 →

Frequently asked questions

How much does Phi-3 Medium cost?

Phi-3 Medium costs $0.1700 per million input tokens and $0.6800 per million output tokens, for a blended reference rate of $0.3230 per million tokens.

What is Phi-3 Medium's context window?

Phi-3 Medium supports up to 128,000 tokens of context in a single request.

What is Phi-3 Medium best for?

Phi-3 Medium is well suited to Single-GPU, Open weights (MIT) and 128K context.

Who makes Phi-3 Medium?

Phi-3 Medium is developed and served by Microsoft. It was released in Apr 2024.

Terms used on this page