Stars
Neural building blocks for speaker diarization: speech activity detection, speaker change detection, overlapped speech detection, speaker embedding
A Minecraft: Java Edition Launcher for Android and iOS based on PojavLauncher. This repository contains source code for Android platform.
Tiny pixel coworkers that live in Paseo and never ask for a raise. A Grok Bot / OpenMausBot alternative plugin for https://paseo.sh/
Run GLM-5.3 (744B), GLM-5.3-Flash (321B), DeepSeek-V4.1-Flash (748B) and MiniMax-M3.1-Flash (428B) on your own GPUs - one engine, one dashboard, OpenAI- and Anthropic-compatible API.
Git designed for AI agents. Open-source, in the spirit of Anthropic's Scion and OpenAI's Pando. Not affiliated with either.
My attempt at getting latest version of the egui demo to work on Android
egui: an easy-to-use immediate mode GUI in Rust that runs on both web and native
GLM-5.3-Flash on two NVIDIA GB10 nodes: versioned ARM64 images, one terminal launcher, native TP2 and image input
Mermaid diagram generation into Claude Code with automatic triggers, MCP coordination
Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.
Blazing-fast LLM inference in pure Rust. No PyTorch and Python runtime.
Rust inference for GLiNER2 checkpoints. Entity/relation extraction, classification, and structured JSON extraction.
Rust reimplementation of Laya, a sub-35ms non-autoregressive typed-decision engine (ModernBERT encoder + RLCD-trained decision head), built from scratch on candle.
Convert Claude Code or Codex into the most intelligent Deep Research Agent. Collect, search, and synthesize web research into a persistent, searchable wiki that builds on itself. Hosted API + MCP: …
A top for local LLMs: memory, paging, GPU and tokens/s for oMLX, Ollama, llama.cpp, vLLM and more
Unofficial native macOS (Apple Silicon) port of the Recoil RTS engine — plays Beyond All Reason. Download the signed, notarized app from Releases →
A script to launch Qwen3.8-Flash-Next in a single DGX-Spark (or its variants) at 40+ tok/s inference speed.
LLM inference in C/C++
Alzeo is an open-source installer that turns an NVIDIA DGX Spark, or any GB10 machine (ASUS, Dell, HP, Lenovo, MSI, GIGABYTE, Acer), into a private AI server in one command: measured model lanes, O…
DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm
Validated recipe for serving Qwen3.6-27B on a single RTX 5090 — full OpenAI API, vision, tool calling, MTP spec-decode
[NeurIPS 2026 Spotlight] Fast, lossless LLM inference via dual-view diffusion decoding.
SNDR Core Engine (Genesis) — vLLM runtime patch-overlay for Qwen3.6 + Gemma4 on consumer NVIDIA (Ampere sm_86, 2× A5000/3090). Qwen3.6-35B-A3B FP8 ~240 tok/s, 27B-int4 hybrid GDN+Mamba, Gemma4 26B/…
Rust for native business logic, Flutter for flexible and beautiful GUI
Qwen3.5-122B-A10B on DGX Spark: 28.3 → 51 tok/s (+80%)
Fullstack app framework for web, desktop, and mobile.