Bhengu.AI.Core
1.1.0
See the version list below for details.
dotnet add package Bhengu.AI.Core --version 1.1.0
NuGet\Install-Package Bhengu.AI.Core -Version 1.1.0
<PackageReference Include="Bhengu.AI.Core" Version="1.1.0" />
<PackageVersion Include="Bhengu.AI.Core" Version="1.1.0" />
<PackageReference Include="Bhengu.AI.Core" />
paket add Bhengu.AI.Core --version 1.1.0
#r "nuget: Bhengu.AI.Core, 1.1.0"
#:package Bhengu.AI.Core@1.1.0
#addin nuget:?package=Bhengu.AI.Core&version=1.1.0
#tool nuget:?package=Bhengu.AI.Core&version=1.1.0
Bhengu.AI.Core
Professional AI model management for .NET applications, with first-class support for Hugging Face models like Microsoft's Phi-3-mini.
Features
✅ Smart Model Downloading
- Resume interrupted downloads
- Checksum verification
- Update detection
🚀 Performance
- Multi-threaded downloads
- Sharded model support
- Progress reporting
🔒 Security
- Authentication support
- File validation
- Secure HTTPS transfers
Installation
dotnet add package Bhengu.AI.Core
Quick Start
using Bhengu.AI.Core;
using Bhengu.AI.Embeddings;
// 1. Initialize downloader (add your HF token if needed)
using var downloader = new HuggingFaceModelDownloader();
// 2. Download Phi-3-mini (automatically caches)
await downloader.DownloadModelAsync(
"microsoft/Phi-3-mini-128k-instruct",
"./ai_models"
);
// 3. Use the model
using var embedder = new Phi3MiniTextEmbedder("./ai_models/microsoft_Phi-3-mini-128k-instruct");
var embedding = await embedder.GenerateAsync("Hello world");
Configuration
Set these optional environment variables:
| Variable | Purpose |
|---|---|
HF_TOKEN |
Hugging Face API token |
HF_CACHE_DIR |
Custom model cache location |
HF_TIMEOUT |
Download timeout in minutes |
Documentation
Full API reference available at:
https://github.com/bhengubv/Bhengu.AI/docs
License
MIT License - Free for commercial and personal use.
| Product | Versions Compatible and additional computed target framework versions. |
|---|---|
| .NET | net9.0 is compatible. net9.0-android was computed. net9.0-browser was computed. net9.0-ios was computed. net9.0-maccatalyst was computed. net9.0-macos was computed. net9.0-tvos was computed. net9.0-windows was computed. net10.0 was computed. net10.0-android was computed. net10.0-browser was computed. net10.0-ios was computed. net10.0-maccatalyst was computed. net10.0-macos was computed. net10.0-tvos was computed. net10.0-windows was computed. |
-
net9.0
- Microsoft.ML.OnnxRuntime (>= 1.16.0)
- SQLitePCLRaw.bundle_green (>= 2.1.0)
- System.Net.Http (>= 4.3.4)
NuGet packages (10)
Showing the top 5 NuGet packages that depend on Bhengu.AI.Core:
| Package | Downloads |
|---|---|
|
Bhengu.AI.Inference
On-device LLM inference for Circle. llama.cpp P/Invoke layer with Qwen 3 chat generator. Native binaries supplied separately (see SETUP.md). |
|
|
Bhengu.AI.Embeddings
On-device text embeddings for Bhengu AI. Sovereign-origin embedder (BGE / Qwen-Embedding) for fully offline semantic search. |
|
|
Bhengu.AI.Memory
On-device episodic memory, affect state, and context window management for Bhengu AI agents. |
|
|
Bhengu.AI.Identity
Cross-device identity for Circle AI. Unified persona key that travels with the person — not the device. Foundation for Bhengu.AI.Sync. |
|
|
Bhengu.AI.Hosting
DI wiring, scheduled tasks, and proactive reasoning host for Bhengu AI agents. |
GitHub repositories
This package is not used by any popular GitHub repositories.