agent.network() Streaming Performance Issues: Batching Instead of Real-time

I'm experiencing a critical streaming performance issue with agent.network() while using the snapshot version 0.0.0-ai-sdk-network-text-delta-20251017172601.

I’m trying to stream responses using agent.network() to my frontend with the AI SDK v5's useChat() hook, similar to how agent.stream() provides character-by-character streaming. My goal is to display hundreds of text-delta chunks in real-time as they arrive.

What’s working:

✅ My backend sends hundreds of text-delta chunks, confirmed via the network tab.
✅ agent.stream() works flawlessly, providing smooth character-by-character streaming.
✅ All chunks are arriving at the browser and can be seen in DevTools.

What’s NOT working:

❌ agent.network() is batching chunks into 2-3 large groups instead of streaming them incrementally.
❌ Text appears all at once or in huge batches, lacking the desired character-by-character flow.
❌ The frontend indicates "2 big chunks" even though the backend is sending over 300 individual chunks.

Current implementation:

const networkStream = await agent.network(content, { /* options */ });  

// Utilizing toAISdkFormat() with manual iteration  
const { toAISdkFormat } = await import("@mastra/ai-sdk");  
const { createUIMessageStream } = await import("ai");  

const uiMessageStream = createUIMessageStream({  
  execute: async ({ writer }) => {  
    for await (const part of toAISdkFormat(networkStream, { from: 'network' })) {  
      writer.write(part); // Chunks are batched here!  
    }  
  }  
});  

return createUIMessageStreamResponse({ stream: uiMessageStream });  

Root cause analysis:

The for await loop is batching the chunks before they are written. In contrast, with agent.stream(), I can simply use:

return streamResult.toUIMessageStreamResponse(); // Works perfectly!  

However, MastraAgentNetworkStream lacks the .toUIMessageStreamResponse() method.

Questions:

1. Will .toUIMessageStreamResponse() be added to MastraAgentNetworkStream in a future snapshot?
2. Is there a method to read from MastraAgentNetworkStream directly without batching using toAISdkFormat()?
3. Should I utilize networkStream.getReader() to read chunks synchronously instead of async iteration?

Environment:

• Mastra snapshot: 0.0.0-ai-sdk-network-text-delta-20251017172601
• AI SDK: v5
• Next.js 15 (App Router)
• Frontend: useChat() hook with streamProtocol: 'data'

The network tab confirms that chunks are being received perfectly, indicating this issue is strictly related to backend transformation and batching. Any guidance would be immensely appreciated! 🙏