move the LLM instance directly to Assistant to make it cleaner to share with tests

package.json

@@ -34,10 +34,10 @@
     "vitest": "^4.1.4"
   },
   "dependencies": {
-    "@livekit/agents": "^1.3.0",
-    "@livekit/agents-plugin-livekit": "^1.3.0",
-    "@livekit/agents-plugin-silero": "^1.3.0",
-    "@livekit/plugins-ai-coustics": "^0.2.10",
+    "@livekit/agents": "^1.4.0",
+    "@livekit/agents-plugin-livekit": "^1.4.0",
+    "@livekit/agents-plugin-silero": "^1.4.0",
+    "@livekit/plugins-ai-coustics": "^0.2.12",
     "dotenv": "^17.4.1",
     "zod": "^3.25.76"
   }

src/agent.test.ts

@@ -1,7 +1,7 @@
 import { dedent, inference, initializeLogger, voice } from '@livekit/agents';
 import dotenv from 'dotenv';
 import { afterEach, beforeEach, describe, it } from 'vitest';
-import { AGENT_MODEL, Agent } from './agent';
+import { Agent } from './agent';
 
 dotenv.config({ path: '.env.local' });
 
@@ -11,20 +11,16 @@ initializeLogger({ pretty: true, level: 'warn' });
 
 describe('agent evaluation', () => {
   let session: voice.AgentSession;
-  let agentLlm: inference.LLM;
   let judgeLlm: inference.LLM;
 
   beforeEach(async () => {
-    agentLlm = new inference.LLM({ model: AGENT_MODEL });
-    // The judge LLM can be a cheaper model since it only evaluates agent responses
     judgeLlm = new inference.LLM({ model: 'openai/gpt-4.1-mini' });
-    session = new voice.AgentSession({ llm: agentLlm });
+    session = new voice.AgentSession();
     await session.start({ agent: new Agent() });
   });
 
   afterEach(async () => {
     await session?.close();
-    await agentLlm?.aclose();
     await judgeLlm?.aclose();
   });
 

src/agent.ts

@@ -1,6 +1,4 @@
-import { dedent, voice } from '@livekit/agents';
-
-export const AGENT_MODEL = 'openai/gpt-5.2-chat-latest';
+import { dedent, inference, voice } from '@livekit/agents';
 
 // Define a custom voice AI assistant by extending the base Agent class
 export class Agent extends voice.Agent {
@@ -40,6 +38,19 @@ export class Agent extends voice.Agent {
         - Protect privacy and minimize sensitive data.
       `,
 
+      // A Large Language Model (LLM) is your agent's brain, processing user input and generating a response
+      // See all available models at https://docs.livekit.io/agents/models/llm/
+      llm: new inference.LLM({ model: 'openai/gpt-5.2-chat-latest' }),
+
+      // To use a realtime model instead of a voice pipeline, replace the LLM
+      // with a RealtimeModel and remove the STT/TTS from the AgentSession
+      // (Note: This is for the OpenAI Realtime API. For other providers, see https://docs.livekit.io/agents/models/realtime/)
+      // 1. Install '@livekit/agents-plugin-openai'
+      // 2. Set OPENAI_API_KEY in .env.local
+      // 3. Add `import * as openai from '@livekit/agents-plugin-openai'` to the top of this file
+      // 4. Replace the llm option with:
+      //    llm: new openai.realtime.RealtimeModel({ voice: 'marin' }),
+
       // To add tools, specify `tools` in the constructor.
       // Here's an example that adds a simple weather tool.
       // You also have to add `import { llm } from '@livekit/agents' and `import { z } from 'zod'` to the top of this file

src/main.ts

@@ -4,7 +4,7 @@ import * as silero from '@livekit/agents-plugin-silero';
 import { audioEnhancement } from '@livekit/plugins-ai-coustics';
 import dotenv from 'dotenv';
 import { fileURLToPath } from 'node:url';
-import { AGENT_MODEL, Agent } from './agent';
+import { Agent } from './agent';
 
 // Load environment variables from a local file.
 // Make sure to set LIVEKIT_URL, LIVEKIT_API_KEY, and LIVEKIT_API_SECRET
@@ -29,12 +29,6 @@ export default defineAgent<ProcessUserData>({
         language: 'multi',
       }),
 
-      // A Large Language Model (LLM) is your agent's brain, processing user input and generating a response
-      // See all providers at https://docs.livekit.io/agents/models/llm/
-      llm: new inference.LLM({
-        model: AGENT_MODEL,
-      }),
-
       // Text-to-speech (TTS) is your agent's voice, turning the LLM's text into speech that the user can hear
       // See all available models as well as voice selections at https://docs.livekit.io/agents/models/tts/
       tts: new inference.TTS({
@@ -52,16 +46,6 @@ export default defineAgent<ProcessUserData>({
       },
     });
 
-    // To use a realtime model instead of a voice pipeline, use the following session setup instead.
-    // (Note: This is for the OpenAI Realtime API. For other providers, see https://docs.livekit.io/agents/models/realtime/))
-    // 1. Install '@livekit/agents-plugin-openai'
-    // 2. Set OPENAI_API_KEY in .env.local
-    // 3. Add import `import * as openai from '@livekit/agents-plugin-openai'` to the top of this file
-    // 4. Use the following session setup instead of the version above
-    // const session = new voice.AgentSession({
-    //   llm: new openai.realtime.RealtimeModel({ voice: 'marin' }),
-    // });
-
     // Start the session, which initializes the voice pipeline and warms up the models
     await session.start({
       agent: new Agent(),