GenAIWiki

Phi-3 Medium

LegacyLatest

Phi-3 Medium is a compact instruct model aimed at strong quality per parameter for on-device and cost-sensitive cloud inference.

Newer version: Phi-4

Provider

Microsoft

Model family

Microsoft Phi

Small LLM

Cost tier

Medium

Status

Legacy

Release May 21, 2024

Why teams choose it

🧠

Complex reasoning

Useful for workflows that require structured thinking, multi-step logic, and deeper analysis than lightweight models provide.

📎

Long-context analysis

Helps teams summarize, compare, and extract insights from long documents without losing important nuance.

⚙️

Microsoft roadmap vigilance

Use published model pages—not stale marketing blurbs—for modalities, quotas, pricing, and policy; schedule revalidation tied to vendor release notes.

✍️

Cost-efficient routing

Useful as part of a routing stack where cheap models handle drafts and confirmations and this tier handles genuinely hard passages.

Tradeoffs to know

  • Context window smaller than frontier chat models—design chunking carefully.
  • Safety: add moderation for user-generated inputs.

When not to use this

  • Self-hosting outcomes depend on hardware, quantization, and ops maturity—budget time beyond swapping an API hostname.
  • May demand more instrumentation than SaaS-managed APIs to duplicate latency, failover, and support guarantees.
  • Benchmark prompts and regressions continuously before rewriting entire routing tables around weights.

Technical specs

Inputs
text
Outputs
text
Capabilities
on-device, reasoning, coding
License
MIT
Model string
phi-3-medium

Benchmarks

No benchmark data yet.

See comparisons →


Microsoft Phi family lineup


Compare with