MODEL · MICROSOFT · 3.8B DENSE
Phi-4 Mini
Microsoft's dense 3.8B instruct model built on synthetic + filtered web data. Punches above its weight on reasoning-heavy prompts in the <5B bracket. MIT license is unusually clean for commercial redistribution.
License: MIT · Context: 128K · Released: February 26, 2025
The decision in five lines
- The call
- Skip for local — for coding
- Best for
- coding · chat · docs · agents
- Runs on
- 23 hardware picks fit (cheapest: Intel Arc B580 12 GB · $249)
- Watch out
- Multilingual depth or vision workloads — Phi-4 Mini is English-heavy and text-only (vision lives in the separate Phi-4-multimodal-instruct).
- Evidence
- Estimated
- 3.8B dense
- PARAMETERS
- DENSE
- TYPE
- 128K
- CONTEXT
- ~3 GB
- VRAM AT Q4
Where we recommend this
Every tier slot in the planner where this model is a top or alternate pick. Pulled live from planner.js — when the planner refreshes, this table stays current.
The call
Microsoft's dense 3.8B instruct model built on synthetic + filtered web data. Punches above its weight on reasoning-heavy prompts in the <5B bracket. MIT license is unusually clean for commercial redistribution.
When not to use: Multilingual depth or vision workloads — Phi-4 Mini is English-heavy and text-only (vision lives in the separate Phi-4-multimodal-instruct).
Runner notes
Ollama tag `phi4-mini`. 200K vocab + GQA means tokenization differs from Llama-family; prompt templates aren't interchangeable.
Hardware that fits
Every hardware pick whose memory fits this model at the quant we recommend. Sorted cheapest-first — the top row is your best-value fit. Click through for the full buyer’s guide.
- Intel Arc B580 12 GBPerfect · 2.7× 12 GB · $249–$299
- NVIDIA RTX 3060 12 GBPerfect · 2.7× 12 GB · $280–$400
- Minisforum UM890 ProPerfect · 5.5× 32 GB DDR5 (shared) · $463–$580 all-in
- RTX 5060 Ti 16 GBPerfect · 3.7× 16 GB · $560–$610
- AMD Radeon RX 9070 XTPerfect · 3.7× 16 GB · $649–$779
- AMD Radeon RX 7900 XTXPerfect · 5.5× 24 GB · $760 used / ~$1,500 new
- Mac Mini M4 16 GBPerfect · 2.5× 16 GB unified · $799 (new floor) / $499–$599 (eBay/residuals)
- NVIDIA RTX 3090 (used, single)Perfect · 5.5× 24 GB · $950–$1,200
- NVIDIA RTX 5070 TiPerfect · 3.7× 16 GB · $980–$1,300
- NVIDIA RTX 5080Perfect · 3.7× 16 GB · $999–$1,400
- MacBook Air M5 24 GBPerfect · 3.7× 24 GB unified · $1,299–$1,699
- Mac Mini M4 Pro 24 GBPerfect · 3.7× 24 GB unified · $1,399
- Dual RTX 3090 (used)Perfect · 11.0× 48 GB · $1,800–$2,500 all-in
- Framework Desktop (Ryzen AI Max+ 395)Perfect · 19.6× 128 GB unified · $1,999–$2,851
- NVIDIA RTX 4090Perfect · 5.5× 24 GB · $2,200–$2,800
- M5 Pro MacBook Pro 48 GBPerfect · 7.4× 48 GB unified · $2,599–$3,099
- NVIDIA RTX 5090Perfect · 7.3× 32 GB · $2,910–$4,300
- Mac Studio M4 Max 64 GBPerfect · 9.8× 64 GB unified · $3,199
- NVIDIA RTX A6000 (48 GB, used)Perfect · 11.0× 48 GB ECC · $3,500–$4,500
- Mac Studio M3 Ultra 96 GBPerfect · 14.7× 96 GB unified · $3,999
- M5 Max MacBook Pro 64 GBPerfect · 9.8× 64 GB unified · $4,499
- NVIDIA DGX SparkPerfect · 19.6× 128 GB unified · $4,699
- Dual RTX 5090Perfect · 14.6× 64 GB (2×32) · $8,500–$10,500
Next step
Find-by-model — see what hardware runs this→