Home/
Part XIII — Expert Mode: Systems, Agents, and Automation/42. Fine-Tuning vs Prompting vs Retrieval (Decision Framework)
42. Fine-Tuning vs Prompting vs Retrieval (Decision Framework)
Overview and links for this section of the guide.
On this page
Decision Framework
When should you prompt, retrieve, or fine-tune? Use this decision tree:
Is the knowledge static and specific to you?
├─ Yes → Does it fit in context?
│ ├─ Yes → Prompting (few-shot, system prompt)
│ └─ No → RAG (retrieve relevant chunks)
└─ No (dynamic/external) → RAG (always fetch latest)
Need specific style/format?
├─ Yes → Can you show examples?
│ ├─ Yes → Few-shot prompting first
│ └─ No → Fine-tuning may help
└─ No → Prompting is sufficientQuick Comparison
| Approach | Best For | Cost | Latency |
|---|---|---|---|
| Prompting | General tasks, quick iteration | Low | Low |
| RAG | Dynamic knowledge, large corpora | Medium | Medium |
| Fine-tuning | Style, format, narrow domains | High | Low |
Where to go next
Explore next
42. Fine-Tuning vs Prompting vs Retrieval (Decision Framework) sub-sections
5 pages