VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning

《开源大模型食用指南》针对中国宝宝量身打造的基于Linux环境快速微调（全参数/Lora）、部署国内外开源大模型（LLM）/多模态大模型（MLLM）教程

See, hear, and interact on your local devices.

Firefly: 大模型训练工具，支持训练Qwen2.5、Qwen2、Yi1.5、Phi-3、Llama3、Gemma、MiniCPM、Yi、Deepseek、Orion、Xverse、Mixtral-8x7B、Zephyr、Mistral、Baichuan2、Llma2、Llama、Qwen、Baichuan、ChatGLM2、InternLM、Ziya2、Vicuna、Bloom等大模型

Cook up amazing AI applications effortlessly with MiniCPM / MiniCPM-V / MiniCPM-o

Easy-to-Use RAG Framework; CCF AIOps International Challenge 2024 Top3 Solution; CCF AIOps 国际挑战赛 2024 季军方案

Windows 本地多模态 AI 桌面桌宠，支持屏幕与音频感知。

Downloadable models and conversion recipes for Apple's Core AI on iPhone and Mac. Chat, vision, speech and generative models with per-model validation records, Swift examples through CoreAIKit, and a downloadable Mac app.

MiniCPM-V apps — fully offline multimodal chat on iOS / Android / HarmonyOS

Inferflow is an efficient and highly configurable inference engine for large language models (LLMs).

Explore LLM model deployment based on AXera's AI chips, It provides OpenAI‑compatible APIs and supports AX620E, AX650 and AX637 series chips.

A custom ComfyUI node for MiniCPM vision-language models, supporting v4, v4.5, and v4 GGUF formats, enabling high-quality image captioning and visual analysis.

Get more intelligence from every bit. Better quantization formats and smarter calibration let larger, stronger models run smoothly on the hardware you already own.

Advanced Android photo editor powered by on-device AI models (LaMa, EdgeSAM, MobileViT) with GPU-accelerated adjustments, voice commands, generative fill, and professional editing tools. Built with Jetpack Compose and ONNX Runtime for fast, offline-first image processing.

A minimal MiniCPM5-1B model inference engine built purely in Rust

CRNN/LPRNet/STNet + CTC + CCPD

为 DeepSeek V4.0 等单模态大模型装上眼睛和耳朵 —— 本地视觉+音频 MCP 服务器，基于 Ollama + MiniCPM-V 4.6 + faster-whisper，图片描述·视频分析·语音转文字

Sydney: Fine-tuned model, companion novel, and philosophical inquiries.

MaskClaw项目构建了一套基于端侧多模态大模型的自进化隐私防护架构，通过视觉脱敏与Agentic 规则演进，实现了 Agent 行为偏好与隐私防护策略的动态协同生长，确保了高危隐私数据在云端交互中的物理级安全与个体化适配。

🤖 Agent Skill 优先：基于原生多模态 MiniCPM v4.7 与可插拔 ASR（faster-whisper / FunASR / 云端 API / MiMo），一句自然语言把视频变成带时间戳字幕、知识文档、图文笔记与 Anki 卡片；全本地全链路闭环，亦附手动 CLI。