vllm

StartLux-Decision

+387

Open decision models from 0.8B to 27B: a calibrated probability for every option in one pass, and a short thought only when unsure. Engine for vLLM, llama.cpp and MLX.

T-Crypt/berth

+100

Decides who gets your GPU. Chat, image and voice engines lease the card in turn instead of grabbing it, so two of them never load at once and crash each other. One Go binary, nothing else to install.

HyperQwen

+187

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.

kvcached

+109

Virtualized Elastic KV Cache for Dynamic GPU Sharing and Beyond

AI-Research-SKILLs

+216

Comprehensive open-source library of AI research and engineering skills for any AI model. Package the skills and your claude code/codex/gemini agent will be an AI research agent with full horsepower. Maintained by Orchestra Research.

EAGLE-4

+66

Official Inference Code of EAGLE-4

verl-omni

+62

Multimodal RL training framework for diffusion & omni models

FunASR

+66

Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.

llama-swap

+57

Reliable model swapping for any local OpenAI/Anthropic compatible server - llama.cpp, vllm, etc