Claude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/MClaude Fable 5$22.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/M
BETA
Subscribe
← Back to Index·All Microsoft models
Microsoft
MicrosoftEfficient

Phi-3.5 MoE

MoE

MoE variant of Phi-3.5 — 16x3.8B experts (~6.6B active). Better quality at similar cost to dense Phi-3.5 mini.

Phi-3.5 MoE is a efficient AI model from Microsoft. It costs $0.160 per million input tokens and $0.640 per million output tokens (blended $0.304/M), with a 128K-token context window.

Released Aug 2024Modalities textOfficial model page ↗API docs ↗Compare with another model →Estimate monthly cost →
INPUT
$0.160/M
per million input tokens
OUTPUT
$0.640/M
per million output tokens
BLENDED 70/30
$0.304/M
default reference rate · how it's calculated →
CONTEXT
128K
128,000 tokens
What it's good at
  • MoE efficiency
  • Open weights
  • 128K context
Typical use cases
  • Self-hosted small chat
  • MoE research
Benchmarks
vs. best public score
MMLU78%
Multitask academic knowledge across 57 subjects.
MATH58%
High-school competition math problems.
HumanEval71%
Python function synthesis from docstrings.
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 →
Phi-4
Frontier SLM · 16K ctx
in $0.070/Mout $0.280/M
Phi-4 Mini
Tiny · 128K ctx
in $0.040/Mout $0.160/M
Phi-3.5 Mini
Compact · 128K ctx
in $0.130/Mout $0.520/M
Phi-3 Medium
Balanced · 128K ctx
in $0.170/Mout $0.680/M
Phi-3 Mini
Tiny · 128K ctx
in $0.130/Mout $0.520/M
Phi-3 Small
Compact · 128K ctx
in $0.150/Mout $0.600/M
Frequently asked questions

How much does Phi-3.5 MoE cost?

Phi-3.5 MoE costs $0.160 per million input tokens and $0.640 per million output tokens, for a blended reference rate of $0.304 per million tokens.

What is Phi-3.5 MoE's context window?

Phi-3.5 MoE supports up to 128K tokens of context (128,000 tokens).

What is Phi-3.5 MoE best for?

Phi-3.5 MoE is well suited to MoE efficiency, Open weights and 128K context.

Who makes Phi-3.5 MoE?

Phi-3.5 MoE is developed and served by Microsoft. It was released in Aug 2024.