The Rise of Small Language Models: Why Llama 3 8B and Phi-3 Rule the Edge

# macOS / Linux
curl -fsSL https://ollama.com/install.sh | sh

# Windows
# Download the installer from ollama.com

ollama run llama3

>>> Why is the sky blue?
The sky appears blue to the human eye because...

async function chatWithLlama(prompt: string) {
  const response = await fetch('http://localhost:11434/api/generate', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
      model: 'llama3',
      prompt: prompt,
      stream: false // Set to true for streaming responses
    })
  });

  const data = await response.json();
  return data.response;
}

// Usage
chatWithLlama("Explain quantum computing like I'm 5")
  .then(console.log);

spring:
  ai:
    ollama:
      base-url: http://localhost:11434
      chat:
        model: llama3
        options:
          temperature: 0.7

import org.springframework.ai.chat.ChatClient;
import org.springframework.web.bind.annotation.*;

@RestController
@RequestMapping("/api/local-ai")
public class LocalAiController {

    private final ChatClient chatClient;

    public LocalAiController(ChatClient chatClient) {
        this.chatClient = chatClient;
    }

    @PostMapping("/generate")
    public Map<String, String> generate(@RequestBody String prompt) {
        // This automatically uses the configured Ollama model
        String response = chatClient.call(prompt);
        return Map.of("response", response);
    }

    // RAG Example with Local Documents
    @PostMapping("/ask-docs")
    public String askDocs(@RequestBody String question) {
        // Assume we have a VectorStore loaded with local docs
        // Spring AI makes switching from Pinecone (Cloud) to
        // a simple in-memory SimpleVectorStore trivial.
        // ... retrieval logic ...
        return "RAG response from local data";
    }
}