Hall of Fame
Previously featured AI tools that are still worth exploring and using
Stable Diffusion WebUI Forge
It was once an enhanced version of the most popular SD WebUI, offering a 60% improvement in memory efficiency compared to the original version.
Last in TOP5: 2025-Q4
Chatterbox (Resemble AI)
An MIT-licensed open-source TTS model built with a Linux-based framework, outperforming ElevenLabs in blind tests.
Last in TOP5: 2026-Q1
Wav2Lip
A classic lip-sync solution that combines any video with any audio to deliver accurate mouth movements, serving as the underlying engine for numerous projects.
Last in TOP5: 2025-Q4
SillyTavern
A visual front-end for immersive Chat & Role-Play experiences, featuring a character card system, a world database, group chat functionality, and Voice & TTS voice synthesis.
Last in TOP5: 2026-Q1
CosyVoice (Alibaba)
It delivers the best results in Chinese—supporting dialects, emotion control, and stream-based inference, with a latency as low as 150ms.
Last in TOP5: 2026-Q1
AnimaHub
An AI-powered system for creating AI Comics — featuring a workflow that includes importing novels, breaking down scripts, creating characters, and synthesizing videos.
Last in TOP5: 2026-Q1
DiffRhythm
This tool enables the creation of a full 4 minutes and 45 seconds song in just 10 seconds, featuring precise alignment with LRC lyrics and outstanding speed.
Last in TOP5: 2026-Q1
Tortoise TTS
High-quality neural text-to-speech that delivers excellent sound quality at the cost of slower speed, making it suitable for offline high-fidelity voice dubbing.
Last in TOP5: 2025-Q3
browser-use
This feature enables the Agent & Automation to operate a real browser, handling form filling, web scraping, and navigation entirely automatically.
Last in TOP5: 2026-Q1
LibreChat
An enterprise-grade multimodal Chat & Role-Play platform featuring multi-user support, RBAC permissions, and audit logs, designed for team deployment via Local Deployment.
Last in TOP5: 2026-Q1
MaxKB
Designed with Chinese users in mind — it supports one-line Local Deployment via Docker and zero-code integration into business systems.
Last in TOP5: 2026-Q1
Notex
The open-source version of NotebookLM supports multiple file formats including PDF, DOCX, and PPTX, along with features such as intelligent Q&A, PPT generation, and mind map creation.
Last in TOP5: 2026-Q1
ChatGPT-on-WeChat (CowAgent)
It was once a benchmark in the Chinese AI Agent community, supporting integration across multiple platforms including WeChat, WeCom, DingTalk, and Lark.
Last in TOP5: 2026-Q1
GFPGAN
A facial restoration algorithm developed by Tencent, serving as a classic tool for restoring old photos and making precise adjustments to faces in AI-generated images.
Last in TOP5: 2025-Q2
AnimateDiff
Bring Motion to SD — Text-to-Video/Image-to-Video with Motion LoRA to control the intensity of movements
Last in TOP5: 2026-Q1
Fish Speech
High-quality multi-language Text-to-Speech — powered by the VITS2 architecture, supporting over 8 languages with audio quality close to that of human speech.
Last in TOP5: 2026-Q1
DeepPrintFilm
Frame-driven AI Comics generation — precise control over keyframes combined with AI-based interpolation for intermediate frames to ensure smooth frame rates.
Last in TOP5: 2026-Q1
InternLM2.5 (Shanghai AI Lab)
Produced by the Shanghai AI Laboratory — featuring strong inference and mathematical capabilities, along with a very long context length, making it the top choice for academic research.
Last in TOP5: 2026-Q1
Rowboat
A powerful solution for meeting scenarios — it connects to Gmail and calendars to build a Knowledge Base & RAG, automatically preparing meeting summaries.
Last in TOP5: 2026-Q1
KrillinAI
A complete solution for video translation and voiceover — featuring AI-powered subtitle translation, synchronized voiceover addition, and one-click publishing, making it an excellent tool for businesses expanding overseas.
Last in TOP5: 2026-Q1
LocalAI
A substitute for the OpenAI API — it runs entirely on CPU and supports multi-model routing, RAG, as well as function calls.
Last in TOP5: 2026-Q2
ChatALL
Ask questions to more than 10 different AIs at the same time and compare their responses side by side — using one-click batch sending to help identify the best answer.
Last in TOP5: 2026-Q2
EchoMimicV2
Developed by Ant Group/DAMO Academy — an audio-driven half-body animation feature that synchronizes the head, gestures, and upper body, making it the top choice for virtual streamers.
Last in TOP5: 2026-Q2
Open Cowork
A desktop-based virtual colleague that allows users to operate a computer just like a human: view the screen, click buttons, fill out forms, export data, and send it to Lark.
Last in TOP5: 2026-Q2
Jan
A stylish and cross-platform desktop AI platform that integrates the MCP protocol along with a rich ecosystem of plugins.
Last in TOP5: 2026-Q1
Cline
A fully self-hosted Agent & Automation extension for VS Code that enables end-to-end execution of tasks such as reading files, writing code, running commands, and browsing the web.
Last in TOP5: 2026-Q1
LangChain + LangGraph
The industry-standard 127K★ LLM application — LangGraph 1.0 official version, featuring a stateful schema-based Agent orchestration engine.
Last in TOP5: 2026-Q1
Sana (NVIDIA)
Created by NVIDIA — 4K high-resolution Image & Video Generation for outstanding image quality.
Last in TOP5: 2026-Q1
Jan (Chat Mode)
The chat & role-play mode of this desktop AI platform features a Local Deployment setup using models combined with the Cortex engine, ensuring zero latency for offline conversations.
Last in TOP5: 2026-Q1
GraphRAG (Microsoft)
Developed by Microsoft — an RAG solution enhanced by knowledge graphs, enabling entity-relation inference along with global understanding capabilities.
Last in TOP5: 2026-Q1
F5-TTS
Ultra-fast cloning in just 2 seconds — featuring the fastest inference speed (RTF of 0.15), with the MIT protocol available for commercial use.
Last in TOP5: 2026-Q1
AIComicBuilder
An AI-powered tool for creating AI Comics — integrating script development, character design, storyboard creation, and Image & Video Generation into a single workflow.
Last in TOP5: 2026-Q1
Yi-1.5 (01.AI)
The best choice for writing long Chinese texts — available in 6B to 34B parameter sizes, with the 34B version offering exceptional value for money.
Last in TOP5: 2026-Q1
LatentSync (ByteDance)
ByteDance – Subspace Diffusion Model for lip synchronization, delivering high-precision mouth shape matching.
Last in TOP5: 2026-Q1
AionUi
A unified desktop interface for various AI tools — it encapsulates the CLI interfaces of models such as Gemini, Claude, and Qwen, along with file management and Excel processing capabilities.
Last in TOP5: 2026-Q1
Langflow
LangChain’s drag-and-drop user interface enables low-code development of RAG and multi-step Agent & Automation workflows.
Last in TOP5: 2025-Q3
Text Generation WebUI (Oobabooga)
A comprehensive full-featured WebUI for developers — supporting LoRA fine-tuning, a rich plugin ecosystem, and compatibility with HuggingFace.
Last in TOP5: 2026-Q1
Roo Code
An AI development team simulator that enables multiple Agent & Automation instances to work in parallel, allowing a single user to oversee and coordinate an entire AI programming team.
Last in TOP5: 2026-Q1
FastGPT
A visual workflow orchestration platform for optimizing Chinese-language Knowledge Base & RAG systems, featuring outbound Chat & Role-Play capabilities as well as integration with WeCom.
Last in TOP5: 2026-Q1
HeyGem
Create a digital avatar in just 30 seconds — available in 8 languages with offline functionality, making it accessible even for complete beginners.
Last in TOP5: 2026-Q1
SongGen
One-step text-to-song conversion — based on a paper presented at ICML 2025, featuring an extremely simplified workflow.
Last in TOP5: 2026-Q1
Kimi K2.5
A comprehensive Office solution that enables lossless conversion between PPT, Word, Excel, and PDF files. It supports up to 100 Agent & Automation instances working in parallel to handle tasks involving as many as 1,500 steps.
Last in TOP5: 2026-Q1
AutoClip
Agent & Automation-powered highlight editing — featuring AI-driven identification of standout moments, automatic scoring, and batch processing, making it an excellent tool for slicing live streams.
Last in TOP5: 2026-Q1
AnythingLLM
A personal Knowledge Base & RAG workspace that enables instant question-answering by dragging in documents, allowing users to set up a private RAG system within just 5 minutes.
Last in TOP5: 2026-Q1
SWE-agent
Developed by Princeton — enables Claude and GPT to independently use terminals to fix GitHub Issues, leading in terms of SWE-bench performance.
Last in TOP5: 2026-Q1
CrewAI
48K★ Chat & Role-Play powered by multi-Agent orchestration — define roles, goals, and backstories, allowing the various Agent & Automation components to work together autonomously.
Last in TOP5: 2026-Q1
CogVideo (Zhipu AI)
Developed by Zhipu AI, this is an optimized Chinese-language video generation model that enables fast creation of 5-second videos.
Last in TOP5: 2026-Q1
big-AGI
An AI programming platform designed for developers, featuring multi-session parallel processing, tool invocation capabilities, code highlighting, and the ability to execute functions.
Last in TOP5: 2026-Q1
OpenSearch + RAG
An AWS open-source search engine combined with an RAG plugin — enabling massive document retrieval, vector search, and full-text search capabilities.
Last in TOP5: 2026-Q1
Diffutoon
Convert real-person videos into anime style — SD-powered video style transfer that transforms content from photo quality to anime quality.
Last in TOP5: 2026-Q1
Baichuan2 (Baichuan AI)
It features robust Chinese language training data—available in 7B and 13B parameter versions—with consistent performance on Chinese instructions, making it suitable for commercial use.
Last in TOP5: 2026-Q1
LiveTalking
A real-time Digital Human end-to-end solution — comprising MuseTalk, Wav2Lip, ER-Nerf, along with WebRTC for live streaming transmission.
Last in TOP5: 2026-Q1
FluxMusic
Rectified Flow Music & Audio Generation: Stable training process combined with fast generation speed
Last in TOP5: 2026-Q1
OfficeClaw (Huawei Cloud)
An enterprise-grade AI-powered Office Productivity Agent & Automation solution featuring a team of multiple Agent & Automation experts, along with fully editable PPT templates for all elements.
Last in TOP5: 2026-Q1
videodl
A video downloader compatible with over 30 platforms — including TikTok, Bilibili, YouTube, and more — that prioritizes high-definition downloads without watermarks.
Last in TOP5: 2026-Q1
KoboldCPP
A specialized inference engine for creative writing — featuring single-file deployment, multimodal capabilities, and MCP bridging.
Last in TOP5: 2026-Q1
Plandex
Designed specifically for large-scale projects — featuring 2 million token context capacity, a differential review sandbox, and support for combining multiple LLMs together.
Last in TOP5: 2026-Q1
AutoGen / AG2 (Microsoft)
48K★ Microsoft’s multi-Agent conversation framework — GroupChat group chat mode, code generation and execution capabilities, and an alternative version of MAF.
Last in TOP5: 2026-Q1
Bernini (ByteDance)
Open-source by ByteDance — a unified framework for Image & Video Generation that offers one-stop solutions for creation, repair, and enhancement.
Last in TOP5: 2026-Q1
Chroma
An AI-native vector database specifically designed for RAG， featuring integration in just 5 lines of code to deliver an optimal development experience for developers.
Last in TOP5: 2026-Q1
MiniCPM-3 (OpenBMB)
Small parameter size, yet impressive capabilities — 4B parameters deliver performance comparable to 7B models, and it runs smoothly on consumer-grade GPUs.
Last in TOP5: 2026-Q1
Sonic (Tencent & Zhejiang Univ.)
High-precision lip synchronization for a single photo — supports long videos with a straightforward workflow.
Last in TOP5: 2026-Q1
SongBloom
A two-phase strategy that starts with a rough approach followed by refinement — alternating between autoregressive sketching and diffusion-based polishing to ensure coherent structure.
Last in TOP5: 2026-Q1
igedits
An open-source alternative to OpusClip — it enables automatic editing of long videos into short clips, featuring functions such as subtitle addition, centered face alignment, and popularity scoring.
Last in TOP5: 2026-Q1
privateGPT
A privacy-first document Q&A system — 100% offline RAG functionality powered by Milvus and Qdrant vector databases.
Last in TOP5: 2026-Q1
Kilo Code
Over 500 pre-trained models, 5 specialized professional modes, and parallel collaboration among multiple Agent & Automation components — the AutoFree mode operates at zero cost.
Last in TOP5: 2026-Q1
Goose
An expandable AI Agent & Automation built with 48K★ Rust — offering integrated installation, execution, editing, and testing capabilities, compatible with any large language model (LLM).
Last in TOP5: 2026-Q1
Hotshot-XL
A native Image & Video Generation model for SDXL — high-quality AI animated GIF creation achieved through fine-tuning of SDXL.
Last in TOP5: 2026-Q1
Agnaistic
A lightweight front-end tool specialized in Chat & Role-Play, featuring Markdown-based character cards, text-based adventure scenarios, and one-click Local Deployment.
Last in TOP5: 2026-Q1
Piper
An extremely lightweight embedded text-to-speech solution that runs on Raspberry Pi, supporting 15 languages with lightning-fast offline performance.
Last in TOP5: 2026-Q1
ComicTranslate
Real-time AR AI Comics Translation — Translate content simply by taking a photo with your smartphone, with support for AR overlay display.
Last in TOP5: 2026-Q1
Phi-3.5 (Microsoft)
Microsoft’s “little powerhouse” — based on the MIT protocol, available in 3.8B/7B/14B parameter versions, offering remarkable performance despite its compact size.
Last in TOP5: 2026-Q1
Ultralight Digital Human
Mobile-based Digital Human – Train and run directly on smartphones, serving as an excellent example of lightweight deployment.
Last in TOP5: 2026-Q1
TangoFlux
Flow Matching for text-to-speech — delivering high-fidelity sound effects, ambient noises, and short musical pieces.
Last in TOP5: 2026-Q1
Lingji Cut
A local-first AI video creation workspace that covers the entire workflow from writing scripts to voiceover recording, shooting, and editing.
Last in TOP5: 2026-Q1