Add basic OpenRouter SDK example extracted from wrapper repo

Author: subtleGradient

typescript/openrouter-sdk/src/basic/example.ts

@@ -0,0 +1,96 @@
+/**
+ * Example usage of the @openrouter/sdk package
+ * 
+ * This demonstrates the OpenRouter TypeScript SDK's idiomatic usage patterns:
+ * - Type-safe client initialization
+ * - Non-streaming chat completions
+ * - Streaming chat completions with async iteration
+ * - Automatic usage tracking
+ * 
+ * Run with: bun examples/basic/example-basic-openrouter-sdk.ts
+ */
+
+import { OpenRouter } from "@openrouter/sdk";
+
+// Initialize the OpenRouter SDK client
+// The SDK automatically reads OPENROUTER_API_KEY from environment
+const openRouter = new OpenRouter({
+  apiKey: process.env.OPENROUTER_API_KEY ?? "",
+});
+
+async function nonStreamingExample() {
+  console.log("=== Non-Streaming Example ===\n");
+
+  // Basic chat completion - returns the full response at once
+  const result = await openRouter.chat.send({
+    model: "openai/gpt-4o-mini",
+    messages: [
+      {
+        role: "user",
+        content: "What is the capital of France?",
+      },
+    ],
+    stream: false, // Explicitly set stream to false for non-streaming
+  });
+
+  // The SDK provides strong typing - result has 'choices' property
+  if ("choices" in result && result.choices[0]) {
+    console.log("Model:", result.model);
+    console.log("Response:", result.choices[0].message.content);
+    console.log("Usage:", result.usage);
+    console.log();
+  }
+}
+
+async function streamingExample() {
+  console.log("=== Streaming Example ===\n");
+
+  // Streaming chat completion - returns an async iterable
+  const stream = await openRouter.chat.send({
+    model: "openai/gpt-4o-mini",
+    messages: [
+      {
+        role: "user",
+        content: "Write a haiku about TypeScript",
+      },
+    ],
+    stream: true, // Enable streaming mode
+    streamOptions: {
+      includeUsage: true, // Include token usage in the final chunk
+    },
+  });
+
+  console.log("Streaming response:");
+  let fullContent = "";
+
+  // The SDK returns an async iterable that you can iterate with for-await-of
+  for await (const chunk of stream) {
+    // Each chunk contains partial data
+    if (chunk.choices?.[0]?.delta?.content) {
+      const content = chunk.choices[0].delta.content;
+      process.stdout.write(content); // Write without newline to see real-time streaming
+      fullContent += content;
+    }
+
+    // Usage stats are included in the final chunk when streamOptions.includeUsage is true
+    if (chunk.usage) {
+      console.log("\n\nStream usage:", chunk.usage);
+    }
+  }
+
+  console.log("\n\nFull response:", fullContent);
+  console.log();
+}
+
+async function main() {
+  try {
+    // Demonstrate both streaming and non-streaming modes
+    await nonStreamingExample();
+    await streamingExample();
+  } catch (error) {
+    console.error("Error:", error);
+    process.exit(1);
+  }
+}
+
+main();