baidu-baige / LoongForge
A high-performance framework for training LLMs, VLMs, diffusion, and embodied models on NVIDIA GPUs and Kunlun XPUs. Supports Pi0.5, GR00T-N1.7, FastWAM, Cosmos3, DreamZero, DeepSeek-V4, GLM-5.3, Kimi-K3 and more.
baidu-baige / LoongForge
FareedKhan-dev / kimi-k3-in-c
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.
MoonshotAI / Kimi-K3
Open Frontier Intelligence
jhd3197 / ServerKit
ServerKit is a lightweight, modern server control panel for managing web applications, databases, and services on your VPS or dedicated server. Built with Python/Flask backend and React frontend.
fcakyon / claude-codex-settings
Battle-tested Claude Code, OpenAI Codex, Cursor configs, plugins, hooks and agents with Kimi, MiniMax and GLM API support.
papersgpt / papersgpt-for-zotero
A powerful Zotero AI and MCP plugin with ChatGPT, Gemini 4, Claude Fable 5.1, Claude Opus 5.5, DeepSeek V4.1, Grok, OpenRouter, Kimi k3, GLM 5.3, SiliconFlow, GPT-oss, Gemma 4, Qwen 3.8
NVIDIA-NeMo / Automodel
🚀 Pytorch Distributed native training library for LLMs/VLMs with OOTB Hugging Face support
PipeNetwork / kimi-k3-mlx
MLX port of moonshotai/Kimi-K3 (2.78T multimodal MoE): streaming converter, REAP expert pruning, and per-language expert-overlap analysis
xorbitsai / inference
Swap GPT for any LLM by changing a single line of code. Xinference lets you run open-source, speech, and multimodal models on cloud, on-prem, or your laptop — all through one unified, production-ready inference API.
gavamedia / deltafin
Run full Kimi K3 on a single device. And an OpenAI-compatible API server for local chat and coding agents.
lucienhuangfu / eLLM
eLLM: Run Long-Horizon Inference Faster on CPUs Than on GPUs
zli12321 / LHTB
Long Horizon Terminal Benchmark with Dense Reward Grading
lordx64 / pentestkit
Autonomous multi-agent pentest framework — plans, exploits, verifies (proof-required), CVSS-scores and writes client-ready reports. 104/104 (100%) on the XBOW validation benchmarks, powered by Kimi K3.
cneuralnetwork / smol-kimi-k3
A 49M-parameter Kimi K3-inspired language model trainable on one 8GB GPU
Neeeophytee / agent-stylebooks
11 installable editorial systems for AI agents, based on leading public style guides.
ciprianveg / gb10-vllm
vLLM inference solutions for NVIDIA GB10 (DGX Spark, sm121) — KIMI-K3 B12X_MLA + DSpark, GLM-5.2
OwenxuSH / shanghaitech-kimi-gateway
Local Kimi-K3 gateway and automatic IDS passkey setup for Kimi Code
VizuaraAI / mini-kimi-k3
The pretraining scaffold behind 'Pretraining a Mini Kimi K3' (free book): six-source HF data mix, benchmark decontamination, tokenizer, WSD training loop, run monitoring. 5B tokens for $252.
ViffyGwaanl / kimi-k3-learn
—
Rurijian / KimiThinkingPrefill
SillyTavern extension: client-side reasoning_content thinking prefill for Kimi/Moonshot models (kimi-k3-jb patch equivalent, no server modifications)