Welcome to Jetson Orin NANO Developer Kit repository
05.Microsoft - Phi-4-mini model
Phi4-mini
Phi4-mini1. Model Size2. Performance3. Using Phi4-mini3.1 Running Phi4-mini3.2 Starting a Conversation3.3 Ending the ConversationReferences
Demo Environment
Development board: Jetson Orin series mainboard
Note: Due to performance limitations, the Jetson Orin Nano 4GB cannot run this model.
Phi-4-mini-instruct is a lightweight open model built on synthetic data and curated public websites, focusing on high-quality, reasoning-dense data.
1. Model Size
| Model | Size |
|---|---|
| phi4-mini:3.8b | 2.5GB |
2. Performance

3. Using Phi4-mini
3.1 Running Phi4-mini
Run the model with the run command; if it is not downloaded yet, Ollama pulls it automatically:
ollama run phi4-mini:3.8b

3.2 Starting a Conversation
xxxxxxxxxxHow many minutes is a quarter of an hour?
Response time depends on the hardware; please be patient.

3.3 Ending the Conversation
Use the Ctrl+d shortcut or /bye to end the conversation!
References
Ollama
Official website: https://ollama.com/
GitHub: https://github.com/ollama/ollama
Phi4-mini
Ollama corresponding model: https://ollama.com/library/phi4-mini