zero-latency

Spout plugin for Unreal Engine 5 using DirectX12

an ultra-low-latency Ethereum indexer running as a Reth ExEx. streams EVM data to ClickHouse with 4ms latency and 71k+ tx/sec throughput.

⚡ HOOX — ultra-low-latency edge trading on Cloudflare Workers. Stack with PYNE + AXIS

A 100% local, zero-latency AI Chrome extension that translates and dubs YouTube videos in real-time using Faster-Whisper, Ollama, and Edge-TTS.

FPS Unlocker & Latency Fix for CS2, Valorant, Apex – No Injections

Optimized for Antigravity AI, Gemini 2026, and High-Performance Agentic Workflows

Universal asynchronous VRAM engine for Ostris AI-Toolkit. Zero-latency compute-transfer overlap with up to 11x training speedup on consumer GPUs.

A highly optimized, thread-safe, zero-latency streaming text interceptor and sensitive words filter for LLM outputs. Supports Python dicts, raw text, and OpenAI SSE streams out-of-the-box. 专为大模型（LLM）流式输出设计的高性能、线程安全、零延迟文本拦截器与敏感词过滤清洗工具。支持原始文本、Python字典、以及 OpenAI SSE 流。

Raspberry Pi smart-home platform that casts precise, zero-latency Adhan audio and HD dashboards to Google Nest Hub, pauses/resumes Android TV media via ADB, and runs a fully local Gemma 3 (Ollama) AI layer for status Q&A and self-diagnosis.

A 4D geometric topology approach to high-speed cryptography.

Zero-frame application buffer - instant GUI application windows with pre-warmed pools

Fix Gemma 4, Llama 4, Qwen 3, Mistral tool-calling failures. Zero-latency regex skill router for Ollama & OpenClaw — 400,000x faster than LLM inference, 100% accurate. No GPU needed.

A high-performance, zero-latency partitioned convolution engine. Designed specifically for real-time guitar cabinet simulation and reverb effects with SIMD optimizations and anti-denormal protection.

High-Level Architectural Framework & Strategic Grant Proposal for Project Catalyst. Open for funding, computational research partnerships, and confidential institutional review.

Glyph is a blazingly fast, minimalist text editor built from the ground up in Rust. It bypasses the bloat of Electron-based editors by utilizing a custom Sovereign Skia Rendering Engine, delivering zero-latency feedback and architectural clarity.

The Hemminki Invariant (ΔI · Φ = W_rec). Universal mathematical framework for post-entropic computing. This identity defines the deterministic transduction of informational state changes (ΔI) into recirculative work (W_rec) using the golden ratio (Φ) as a phase-harmonic constant to eliminate thermal dissipation with zero-latency state recovery.

Predictive preemption engine monitoring active UI and audio telemetry at 10Hz to forecast LLM prompts. Speculatively streams KV cache snapshots to local inference servers while dynamically throttling background workloads via CUDA MPS or OS level process priority.

This repository mimics a live distributed application environment with externalized configurations. Achieved server update with zero downtime, along with version control of the configurations.

Zero-added-latency parametric EQ + dynamic EQ (external sidechain) + band-isolated harmonic generation + compressor VST3/AU plugin (JUCE/C++), with ballistics-accurate metering and presets. AI-assisted, human-directed. Downloads: https://stoatworks-labs.com/software/zero-eq/ | Try it in-browser: https://zero-eq-sim.allan-sargeant.workers.dev

Zero-latency, privacy-first AI English speaking friend desktop app. Features local STT/TTS, direct RAM audio processing, and asynchronous BYOK LLM streaming.