Question 1

How to use LLM Tornado with local models like Ollama?

Accepted Answer

Install the base package, create a TornadoApi instance with the local Ollama URI, and use Chat.CreateConversation with a model name string. Tornado handles request transformations automatically, as shown in the self-hosted demo.

Question 2

LLM Tornado vs Semantic Kernel: which is better for .NET AI?

Accepted Answer

LLM Tornado excels in multi-provider support and agent orchestration, while Semantic Kernel is more integrated with Microsoft's ecosystem. Tornado can interoperate via Microsoft.Extensions.AI, but choose based on need for provider flexibility versus Microsoft tooling.

Question 3

Does LLM Tornado support Azure OpenAI Service?

Accepted Answer

Yes, Azure is listed as a built-in connector. Configure it using API keys and endpoints similar to other providers, with support for Azure-specific features via VendorExtensions.

Question 4

How to handle errors and retries in LLM Tornado agents?

Accepted Answer

Use the GetResponseRichSafe API for network-level error handling, or implement try-catch blocks. The SDK includes features like zig-zag model fallback in ToolkitChat for resilience against provider downtime.

Question 5

Can I use LLM Tornado in a Blazor WebAssembly app?

Accepted Answer

Yes, it's a .NET library compatible with Blazor WASM, but API calls run client-side, so manage CORS and security for external providers. Examples like Flowbite Blazor integration demonstrate this use.

Question 6

What are the performance costs of LLM Tornado's abstractions?

Accepted Answer

The SDK is optimized for performance with over 500 tests, but high-level abstractions may add minimal overhead compared to direct API calls. The trade-off is reduced boilerplate and improved maintainability for complex workflows.

LlmTornado

What is LlmTornado?

Overview

Use Cases

Best For

Related Projects

Found a gem we're missing?

Not Ideal For

Pros & Cons

Pros

Cons

Frequently Asked Questions