Search

Make videos on the computer you already own. FreeVideo runs MiniMax H3 in as little as 8 GB of VRAM and 16 GB of RAM, and adapts its acceleration path to your hardware.

Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent, 150k-262k context. vLLM patches, requant pipeline, benchmarks.

Xolotl: run DeepSeek-V4.1-Flash, all 552B, as a local coding agent on one gaming PC (RTX 4090 + 110 GB RAM): ~33 tok/s on agent tasks, ~41 in fast mode, 512K context. Plus GLM-5.3-Flash, Qwen3.8 and LFM3-Omni. MoE experts tiered across VRAM, pinned RAM and NVMe, with Xolotl Direct NVMe->VRAM. OpenAI + Anthropic API.

Run 35B MoE models in ~2.5 GB of RAM. Weights stream from SSD; a trained router predicts loads a step ahead. Mac, iPhone, Android, Windows supported

Implement one small API on SQLite in any language; we benchmark it on a $12 server (1 vCPU, 2 GB)

Gemma 4 26B-A4B inference in ~2 GB of RAM on any M-series MacBook

Optimized SGLang runtime for Qwen3.8-27B FP8 with DFlash2 and Qwen3.8 Flash-Next NVFP4 with FR-Spec on one NVIDIA RTX PRO 6000 Blackwell 96 GB GPU (SM120): 524K context, HiCache and NIXL.

Pure-Rust inference engine for Qwen GGUF models (Qwen3 to Qwen3.8, dense and MoE up to 770B+). CUDA and Vulkan (NVIDIA, AMD), hybrid CPU+GPU experts, MTP speculative decoding, OpenAI and Anthropic compatible server. Up to ~1.7x llama.cpp decode on a 12 GB RTX 3060.

WLA DX - Yet Another ANSI C GB-Z80/Z80/Z80N/6502/65C02/65CE02/65816/68000/6800/6801/6809/8008/8080/HUC6280/SPC-700/SuperFX/Cx4/SH-2 Multi Platform Cross Assembler Package

中国行政区划代码【省、市、区县、乡镇街道】中国省市区镇三级四级联动行政区划数据（GB/T 2260）

A biblatex implementation of the GB/T7714-2015 bibliography style || GB/T 7714-2015 参考文献著录和标注的biblatex样式包

Game Boy: Complete Technical Reference

A quick and easy to use drag and drop retro game creator for your favourite handheld video game system

Learn to create games for the Game Boy using Assembly

Assembled Genomes Compressor (AGC) squeezing 290 GB of Human Pangenome Project assemblies down to <1.5 GB (>190× ratio), with sample and contig extraction in seconds.

The Fastest Unified Streaming SQL Engine in a Single C++ Binary. ⚡ Millisecond latency. 100+ GB/s throughput. Continuously compute real-time context from streams, logs, metrics, events, and CDC.

纯协议层框架，屏蔽 SIP 协议细节。业务方实现 Listener 接口（推荐）或监听 Layer 1 协议事件 接收消息，通过 ClientCommandSender / ServerCommandSender 发送命令，不直接接触 JAIN-SIP。框架内置 GB28181-2016 + GB/T 28181-2022 协议实现，单 JVM 可同时启用平台服务端（gb28181-server）和设备客户端（gb28181-client），支持级联代理场景。

This is a set of xml enabling custom systems like Switch, hacks of gb,gbc,gba,nes, snes and gen

Language model tokenization at GB/s

Static recompiler for Game Boy and Game Boy Color ROMs that emits portable C projects with an SDL2 runtime.

Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.

Run MoE models bigger than your RAM. Frontier-size MoE on a 12 GB phone, CPU only, lossless, on stock llama.cpp

TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM

A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.

Run a 105 GB AI model on a Mac that can't hold it. Slotstream streams Qwen3.8-Flash-Next (125B mixture of experts) from your SSD and caches the busiest experts in memory, so it runs on Macs with 16 to 64 GB. One native Swift binary on MLX and Metal, no Python, offline. Works with Claude Code, Codex and Ollama or OpenAI clients.

Qwen3.8-27B on 16-32 GB Nvidia GPUs one-click install for Windows / Linux

跨平台中文公文排版 Skill，生成参照 GB/T 9704-2012 的可编辑 DOCX

37 tok/s from Qwen3-Next-80B (48 GB) on a 32 GB laptop, 24 without drafts vs 2.8 for llama.cpp; reading 62 vs 16 tok/s. Faster than llama.cpp on all 13 models compared in Q4_K_M. LLMs bigger than your RAM, pure C, CPU or GPU.

Strata fork: Qwen3.8-Flash-Next 125B MoE in NVFP4 on one RTX 20-50 card (12 GB+) and 64 GB of RAM or more. Setup installs our GPTQ quants (huihui abliterated, OrcaRouter) from Hugging Face. CPU experts on AVX-512 or AVX2 (+AVX-VNNI), all experts in RAM from 92 GB, a low-RAM mode below, vision, W4A8 prompts. Windows / CUDA 13.

GLM-5.3-Flash EXL3 on one 24 GB RTX 3090 + DDR4: elastic GPU expert cache, zero-copy experts, AVX2 CPU tier, OpenAI API