Skip to content

@defilantech Defilan Technologies

All

    Repositories list

    13 repositories

    • llmkube-bench

      Public
      Reproducible llama.cpp vs vLLM benchmark on Kubernetes for local LLM inference. Manifests, load harness, and full bake-off results for Qwen3.6-27B on 2x RTX 506…
      kubernetesbenchmarkgpu
      kubernetesbenchmarkgpuinferencellmllama-cppvllmlocal-llm
      Python
      •
      Apache License 2.0
      •2•5•1•4•Updated Oct 1, 2026Oct 1, 2026

    • llmkube-runtimes

      Public
      LLMKube inference runtime images (AMD/Vulkan first). Build-from-source, hardware-gated CI.
      Python
      •
      Apache License 2.0
      •3•1•3•4•Updated Oct 1, 2026Oct 1, 2026

    • LLMKube

      Public
      Kubernetes operator for self-hosted LLM inference across a heterogeneous GPU fleet: NVIDIA CUDA, AMD Vulkan, and Apple Silicon Metal. Runtimes: llama.cpp, vLLM,…
      kubernetesaimetal
      kubernetesaimetalgpuinferenceself-hostednvidiakubernetes-operatorautoscalinghomelab
      Go
      •
      Apache License 2.0
      •35•226•98•7•Updated Sep 30, 2026Sep 30, 2026

    • homebrew-tap

      Public
      Homebrew tap for LLMKube
      Ruby
      •0•0•0•0•Updated Sep 30, 2026Sep 30, 2026

    • occonfig

      Public
      Named model profiles for opencode configuration
      rustcliopencode
      rustcliopencodedeveloper-tools
      Rust
      •
      Apache License 2.0
      •0•0•0•7•Updated Sep 28, 2026Sep 28, 2026

    • infercost

      Public
      Kubernetes-native cost intelligence for on-premises AI inference. Computes true cost-per-token from GPU amortization, electricity, and real power draw.
      kubernetesopen-sourceai
      kubernetesopen-sourceaigpuinferenceprometheuscostkubernetes-operatorfinopsllm
      Go
      •
      Apache License 2.0
      •2•6•8•6•Updated Sep 14, 2026Sep 14, 2026

    • llmkube-kueue

      Public
      Kueue integration for LLMKube: external jobframework component that lets Kueue admit, queue, and account GPU quota for InferenceServices
      Go
      •
      Apache License 2.0
      •0•1•2•0•Updated Jul 24, 2026Jul 24, 2026

    • mlx-server

      Public
      OpenAI-compatible HTTP server for mlx-swift-lm on Apple Silicon. Drop-in replacement for llama-server with state-of-the-art speculative decoding.
      Swift
      •
      Apache License 2.0
      •1•1•1•2•Updated Jun 28, 2026Jun 28, 2026

    • foreman-python-example

      Public
      Minimal Python project demonstrating LLMKube's Foreman agentic-coding harness on a non-Go codebase (language-configurable gate).
      Python
      •
      Apache License 2.0
      •1•0•0•0•Updated Jun 25, 2026Jun 25, 2026

    • llmkube-bootstrap

      Public
      One-command setup for a fresh Apple Silicon Mac to run local LLM inference on LLMKube.
      Shell
      •
      Other
      •0•0•1•0•Updated May 24, 2026May 24, 2026

    • vecsmith

      Public
      Open-source text-to-SVG generation pipeline. Turn text prompts into production-ready vector graphics using LLM prompt enhancement, Flux image generation, and vt…
      pythonsvgflux
      pythonsvgfluxvector-graphicstext-to-imagefastapigenerative-ai
      Python
      •
      Apache License 2.0
      •1•2•0•1•Updated Mar 24, 2026Mar 24, 2026

    • .github

      Public
      0•0•0•0•Updated Mar 23, 2026Mar 23, 2026

    • issueparser

      Public
      LLM-powered GitHub issue theme analyzer. Scan repositories for issues and use AI to identify common pain points, recurring themes, and actionable insights.
      githubkubernetesgolang
      githubkubernetesgolangaianalyticsissue-trackingllmllmkube
      Go
      •
      Apache License 2.0
      •0•0•0•0•Updated Nov 29, 2025Nov 29, 2025
    ProTip! When viewing an organization's repositories, you can use the props. filter to filter by custom property.