The easiest way to use Ollama in .NET
Run, fine-tune and serve real LLMs entirely in-process in .NET — on the CPU, on-device, offline. No Python, no GPU, no cloud.
Security monitoring middleware for IChatClient (Microsoft.Extensions.AI). 55 detectors for prompt injection, hallucination, PII leakage, and operational anomalies. Intervention engine, embedded dashboard, audit forwarders to Azure Sentinel + OpenTelemetry. Drop-in middleware for any LLM client.
Composable Microsoft.Extensions.AI IChatClient middlewares for .NET — tool-call tracking with in-band streaming progress, per-tool token usage reports, and satellite packages for MCP tools and Microsoft Agent Framework agents-as-tools.
A minimalistic but extensible chat control library for Windows Forms
Use browser-provided, on-device language models from Blazor through Microsoft.Extensions.AI and the Prompt API.
Minimal .NET 10 console samples calling Azure OpenAI via Microsoft.Extensions.AI IChatClient, including function/tool calling and a file-based app.
A local .NET 10 spike that reflectively optimizes an IChatClient overlay across four axes — instructions, tool descriptions, model selection, and skills — using Microsoft.Extensions.AI.Evaluation. Inspired by Azure AI Foundry's Agent Optimizer.
A 60-second check that your .NET AI dev environment works - local (Ollama) + cloud (GitHub Models) via one IChatClient. Companion to the AI-Powered APIs in .NET course.
To associate your repository with the ichatclient topic, visit your repo's landing page and select "manage topics."