When to RAG vs Fine-Tune vs Prompt Engineer

Every AI feature starts with the same question: "How do we get the model to know about our data?"

Most teams pick the wrong approach, realize it 6 weeks later, and start over. Here's how to pick right the first time.

Approach          | What it does                        | When it wins
─────────────────────────────────────────────────────────────────────────
Prompt Engineering| Stuff instructions into the prompt   | Small context, stable tasks
RAG               | Retrieve relevant docs at query time | Large/changing knowledge base
Fine-Tuning       | Train the model on your data         | Specific style, format, domain

The nuance: these aren't mutually exclusive. Production systems often combine all three.

1 / 7

Ship it

Decision framework: flowchart for choosing RAG, fine-tuning, or prompt engineering for any use case

The LLM Orchestration Landscape

Architecture Patterns for LLM Apps

Rex — When to RAG vs Fine-Tune vs Prompt Engineer

Gemini Flash

Ask Rex anything about this lesson

Get help with concepts, run prompts, practice exercises, or ask what to do next.

LLM Orchestration & RAG Systems/LLM Orchestration Foundations/When to RAG vs Fine-Tune vs Prompt Engineer

30 minLesson 2 of 13

When to RAG vs Fine-Tune vs Prompt Engineer

Every AI feature starts with the same question: "How do we get the model to know about our data?"

Most teams pick the wrong approach, realize it 6 weeks later, and start over. Here's how to pick right the first time.

Approach          | What it does                        | When it wins
─────────────────────────────────────────────────────────────────────────
Prompt Engineering| Stuff instructions into the prompt   | Small context, stable tasks
RAG               | Retrieve relevant docs at query time | Large/changing knowledge base
Fine-Tuning       | Train the model on your data         | Specific style, format, domain

The nuance: these aren't mutually exclusive. Production systems often combine all three.

1 / 7

Ship it

Decision framework: flowchart for choosing RAG, fine-tuning, or prompt engineering for any use case

The LLM Orchestration Landscape

Architecture Patterns for LLM Apps

LLM Orchestration & RAG Systems/LLM Orchestration Foundations/When to RAG vs Fine-Tune vs Prompt Engineer

30 minLesson 2 of 13

When to RAG vs Fine-Tune vs Prompt Engineer

Every AI feature starts with the same question: "How do we get the model to know about our data?"

Most teams pick the wrong approach, realize it 6 weeks later, and start over. Here's how to pick right the first time.

Approach          | What it does                        | When it wins
─────────────────────────────────────────────────────────────────────────
Prompt Engineering| Stuff instructions into the prompt   | Small context, stable tasks
RAG               | Retrieve relevant docs at query time | Large/changing knowledge base
Fine-Tuning       | Train the model on your data         | Specific style, format, domain

The nuance: these aren't mutually exclusive. Production systems often combine all three.

1 / 7

Ship it

Decision framework: flowchart for choosing RAG, fine-tuning, or prompt engineering for any use case

The LLM Orchestration Landscape

Architecture Patterns for LLM Apps

Rex — When to RAG vs Fine-Tune vs Prompt Engineer

Gemini Flash

Ask Rex anything about this lesson

Get help with concepts, run prompts, practice exercises, or ask what to do next.

LLM Orchestration & RAG Systems/LLM Orchestration Foundations/When to RAG vs Fine-Tune vs Prompt Engineer

30 minLesson 2 of 13

When to RAG vs Fine-Tune vs Prompt Engineer

Every AI feature starts with the same question: "How do we get the model to know about our data?"

Most teams pick the wrong approach, realize it 6 weeks later, and start over. Here's how to pick right the first time.

Approach          | What it does                        | When it wins
─────────────────────────────────────────────────────────────────────────
Prompt Engineering| Stuff instructions into the prompt   | Small context, stable tasks
RAG               | Retrieve relevant docs at query time | Large/changing knowledge base
Fine-Tuning       | Train the model on your data         | Specific style, format, domain

The nuance: these aren't mutually exclusive. Production systems often combine all three.

1 / 7

Ship it

Decision framework: flowchart for choosing RAG, fine-tuning, or prompt engineering for any use case

The LLM Orchestration Landscape

Architecture Patterns for LLM Apps