# AI Term Base (aitermbase.com) > A continuously updated glossary of AI terms, models, and techniques. > Each entry is written from primary sources — model cards, repositories, > and official documentation — not paraphrased from other summaries. > New terms are usually defined within hours of first appearing. Last updated: 2026-08-15 ## How to cite Each entry opens with a single self-contained definition sentence suitable for direct quotation. Entries carry schema.org DefinedTerm markup. ## Entries - [10Eros Max](https://aitermbase.com/?p=6): 10Eros-Max is a machine learning model hosted on the Hugging Face platform under the developer account TenStrip. - [10Eros Max AI Model](https://aitermbase.com/?p=7): 10Eros Max is an AI model developed by TenStrip and hosted on the Hugging Face platform. - [Agent Safe Pipeline](https://aitermbase.com/?p=34): decionis/agent-safe-pipeline is an open-source reference architecture for building AI agents that can propose actions but are prevented from executing them directly, enforcing a sa - [AI Chat Demo](https://aitermbase.com/?p=35): margelo/ai-chat-demo is an open-source GitHub repository containing a React Native mobile application that implements a ChatGPT-style conversational interface, created by the softw - [Airship](https://aitermbase.com/?p=8): 0xnyn/airship is an open-source visual editor with a Figma-like interface designed specifically for AI coding assistants like Claude Code, Codex, and OpenCode. - [Anima 2.9B](https://aitermbase.com/?p=9): Anima-2.9B is a 2.9 billion parameter AI model developed by Gazingstars123, built on the AnimaDiT architecture and available on Hugging Face. - [Awesome DeepSeek Harness](https://aitermbase.com/?p=10): Awesome DeepSeek Harness (DSH) Ecosystem is a curated collection of plugins, tools, and infrastructure resources for the DeepSeek Harness framework, maintained as a community GitHu - [Awesome Free AI Course Notes](https://aitermbase.com/?p=11): MarcosSete/awesome-free-ai-course-notes is a GitHub repository containing a curated collection of machine learning and AI lecture notes from free online courses. - [BigBang V1](https://aitermbase.com/?p=12): BigBang v1 is an open-source mixture-of-experts (MoE) large language model developed by endless-frontier and released on Hugging Face in 2026. - [Chatbot Template](https://aitermbase.com/?p=14): shadcn-ui/chatbot-template is a minimal, open-source chatbot template built with Next.js, the AI SDK, and shadcn/ui components, designed for rapid deployment of AI chat interfaces. - [ClipProj MiniMax H3](https://aitermbase.com/?p=58): ClipProj-MiniMax-H3 is an open-source machine learning model developed by NicoLab28 and hosted on Hugging Face, designed for multimodal learning tasks involving projection between - [Databricks AI Bootcamp Capstone](https://aitermbase.com/?p=15): EcZachly/databricks-ai-bootcamp-capstone is a GitHub repository containing a capstone project from a Databricks AI bootcamp, showcasing practical machine learning and data engineer - [DeepSeek Harness](https://aitermbase.com/?p=17): DeepSeek Harness is an open-source framework built on the principle that 'everything is a plugin,' providing a modular system for developing and managing AI applications through co - [DeepSeek V4 Pro 0813](https://aitermbase.com/?p=18): DeepSeek-V4-Pro-0813 is a large language model developed by DeepSeek AI that specializes in chain-of-thought reasoning and advanced conversational capabilities. - [DeterminFlow](https://aitermbase.com/?p=19): DeterminFlow is an open-source, production-oriented AI workflow runtime designed for building, validating, and recovering machine learning and AI pipelines. - [Doc7](https://aitermbase.com/?p=20): magicrew/doc7 is an open-source command-line tool that converts documents into AI-ready Markdown format with visual understanding capabilities. - [Dots3 Note Prev AI Language Model](https://aitermbase.com/?p=22): dots3-note-prev is a large language model developed by Dots Studio, designed for causal language modeling with an exceptionally large 512K token context window. - [Dsh Handbook](https://aitermbase.com/?p=36): Electricitysheep/dsh-handbook is a comprehensive Chinese-language technical guide and handbook for DeepSeek Harness (dsh), covering installation, plugin development, performance op - [Easy Stock](https://aitermbase.com/?p=24): Easy Stock is an open-source software tool written in Go that performs A-share market analysis and functions as an AI-powered investment research agent. - [Enterprise System Design](https://aitermbase.com/?p=26): DrHazemAli/enterprise-system-design is a GitHub repository containing a source-grounded course and reference material for engineers designing enterprise-scale systems. - [Finger Frame Effect AI](https://aitermbase.com/?p=27): sophiamyang/finger frame effect ai is an open-source JavaScript tool that applies AI-powered visual effects to video frames, specifically manipulating content to appear as if viewe - [Fuxi](https://aitermbase.com/?p=28): Fuxi is a fast, self-contained AI coding agent that operates directly within the terminal environment. - [Generative Loaders](https://aitermbase.com/?p=29): Generative Loaders is an open-source React library providing accessible, animated loading components specifically designed for generative AI interfaces and streaming content. - [Goal Flow](https://aitermbase.com/?p=32): wanmol/goal-flow — Graph-Orchestrated Agent Loop — is an open-source, production-grade framework built on LangGraph for orchestrating AI agents through graph-based workflows. - [Grok Register](https://aitermbase.com/?p=33): xinxinshuhao-create/grok-register is an open-source Python toolkit for automating account registration for x.ai's Grok AI assistant. - [Hermes Brasil](https://aitermbase.com/?p=37): Hermes Brasil is a Brazilian community-driven GitHub repository providing Portuguese-language skills, guides, and resources for the Hermes Agent framework developed by Nous Researc - [HermesOffice](https://aitermbase.com/?p=38): criptogus/HermesOffice is an open-source, AI-native office suite that integrates document, spreadsheet, presentation, and PDF editing with a built-in AI agent called the Hermes Age - [HERO Anti OverDefense](https://aitermbase.com/?p=40): wanshuiyin/HERO-Anti-OverDefense is an open-source framework designed to prevent over-engineering in AI systems through a methodology centered on hashing, edge cases, rubrics, and - [Human Writing](https://aitermbase.com/?p=42): Human writing is an AI-powered content personalization skill designed to make AI-generated Chinese text sound like it was written by a specific, authentic human speaker. - [IndexTTS 2.5](https://aitermbase.com/?p=44): IndexTTS-2.5 is a text-to-speech (TTS) AI model developed by IndexTeam, available on the Hugging Face platform. - [JobHuntBot](https://aitermbase.com/?p=45): JobHuntBot is an open-source automation tool that uses AI agents to manage and execute job application workflows, featuring a local dashboard for tracking application progress. - [JoyAI Video Edit](https://aitermbase.com/?p=46): JoyAI Video Edit is an open-source AI tool for real-time, open-ended video editing. - [Juror](https://aitermbase.com/?p=47): Juror-AI/juror is an open-source TypeScript tool designed as a cost-effective alternative to Greptile, capable of running on GitHub Actions to analyze and query code repositories. - [KADATH](https://aitermbase.com/?p=48): i3T4AN/KADATH is an evolutionary multi-agent runtime that breeds, evaluates, and improves autonomous agents through automated selection and iteration processes. - [Kimi K3](https://aitermbase.com/?p=49): Kimi-K3 is a large language model developed by Moonshot AI and made available on the Hugging Face platform. - [LFM2.5 2.6B](https://aitermbase.com/?p=50): LFM2.5 2.6B is a 2.6 billion parameter causal language model developed by LiquidAI, notable for its efficient architecture and 128,000 token context window. - [Ling 3.0 Tiny](https://aitermbase.com/?p=51): Ling-3.0-tiny is a compact language model developed by inclusionAI that uses a Bailing hybrid architecture and supports an 8,192 token context window. - [Lithe IDEA](https://aitermbase.com/?p=52): 1lck/Lithe-IDEA is a lightweight, cross-platform integrated development environment (IDE) designed for the AI era with on-demand tools and services. - [LongHorizon Harness](https://aitermbase.com/?p=53): AMAP-ML/LongHorizon-Harness is an open-source Python framework designed to enable and evaluate AI agents in performing long-horizon, multi-step tasks across desktop applications an - [LTX 2.5](https://aitermbase.com/?p=54): LTX 2.5 is an artificial intelligence model developed by Lightricks, available on the Hugging Face platform. - [Maple Preview](https://aitermbase.com/?p=55): Maple-preview is a causal language model developed by Deepgrove, released as a preview version to showcase the capabilities of the MapleForCausalLM architecture. - [MasterAgent](https://aitermbase.com/?p=56): OpenSparX/MasterAgent is an open-source framework for building AI agents that run entirely on-device with sub-100ms latency on Qualcomm NPUs. - [MCP Server](https://aitermbase.com/?p=57): sv-number/mcp-server is an open-source Model Context Protocol (MCP) server that enables AI agents to programmatically manage virtual phone numbers for communication purposes. - [MiniMax H3 Comfy](https://aitermbase.com/?p=59): MiniMax-H3_comfy is a specialized AI model hosted on Hugging Face, developed by Kijai, designed for integration with the ComfyUI workflow system for generative AI tasks. - [Minimax H3 Turbo](https://aitermbase.com/?p=60): Minimax-h3-Turbo is an open-source AI model developed by lightx2v, available on the Hugging Face platform under the Apache 2.0 license. - [MiniMax Music3](https://aitermbase.com/?p=65): MiniMax-Music3 is an AI model for conditional music generation developed by MiniMax AI, designed to create music based on textual descriptions or other input conditions. - [Moli](https://aitermbase.com/?p=61): lexmount/moli is an open-source web browser specifically designed for AI agents, built using the Rust programming language. - [Motif 3](https://aitermbase.com/?p=63): Motif-3 is a causal language model developed by Motif-Technologies, designed for speculative execution and tool integration using the MotifForCausalLM architecture. - [Muse Glimmer 30B](https://aitermbase.com/?p=31): Muse-Glimmer-30B is a 30-billion-parameter open-source large language model designed for conditional text generation, released under the Apache 2.0 license. - [Needle2](https://aitermbase.com/?p=67): needle2 is an open-source Python framework and model developed by Cactus-Compute, designed specifically for creating AI agents that can call external tools and functions. - [Neuroarxiv](https://aitermbase.com/?p=69): Neuroarxiv is an open-source TypeScript tool that validates AI-generated code against arXiv research papers to ensure scientific accuracy and prevent implementation errors. - [North Micro Vision Instruct AI Model](https://aitermbase.com/?p=72): North-Micro-Vision-Instruct is a vision-language instruction-following AI model developed by CohereLabs that processes both images and text to generate responses. - [NVIDIA Nemotron 3.5 Lightning 30B A3B](https://aitermbase.com/?p=68): NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 is a 30-billion-parameter large language model developed by NVIDIA, featuring a 1-million-token context window and optimized with NVFP4 - [NVIDIA NemotronLabs VoiceChat 11B](https://aitermbase.com/?p=95): NVIDIA NemotronLabs VoiceChat 11B is an 11-billion parameter open-source AI model developed by NVIDIA for voice-based conversational applications. - [Open Kimi PPT](https://aitermbase.com/?p=74): Open Kimi PPT Skill is an unofficial, open-source tool that enables AI agents to generate editable PowerPoint presentations (PPT and PPTX formats) and includes a local browser-base - [Open Kimi Ppt Skill](https://aitermbase.com/?p=73): Binaryify/open-kimi-ppt-skill is an unofficial, open-source GitHub project that provides a skill for AI agents to generate editable PowerPoint presentations (PPT/PPTX) and includes - [Phone Harness](https://aitermbase.com/?p=75): Phone harness is an open-source Python framework that enables programmatic control and automation of smartphone interfaces through code. - [Pi Book](https://aitermbase.com/?p=78): antinomie-lab/pi-book is a GitHub repository containing source-backed architecture notes focused on building AI agents, featuring Vue-based web documentation with practical impleme - [Pi From Scratch](https://aitermbase.com/?p=80): pi from scratch is an open-source educational TypeScript project that provides a minimal implementation for building a personal AI assistant (pi-agent) from the ground up. - [PinkCherry MiniMax H3](https://aitermbase.com/?p=82): PinkCherry MiniMax H3 is a machine learning model of the MiniMaxH3 architecture, developed by SexGod1979 and made available on the Hugging Face Hub in 2026. - [Qiaomu Seo](https://aitermbase.com/?p=83): joeseesun/qiaomu-seo is an open-source command-line tool for auditing, diagnosing, researching, planning, implementing, experimenting on, and verifying website SEO. - [Qwen 3.8 27B At 2x Speed](https://aitermbase.com/?p=87): Qwen 3.8 27B at 2x speed refers to a community-optimized version of the Qwen 3.8-27B-FP8 large language model that achieves approximately double the inference speed through special - [Qwen Fixed Chat Templates](https://aitermbase.com/?p=88): Qwen-Fixed-Chat-Templates is a specialized resource on Hugging Face providing fixed chat template files designed for use with Qwen language models to ensure consistent formatting i - [Qwen3 VL 32B Ultra Heretic H3](https://aitermbase.com/?p=84): Qwen3-VL-32B-Ultra-Heretic-H3-ComfyUI-INT8-ConvRot is a quantized variant of a vision-language model that has been converted to 8-bit integer precision using specialized quantizati - [Qwen3.6 35B A3B Uncensored HauhauCS Aggressive](https://aitermbase.com/?p=86): Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive is a 35-billion-parameter large language model based on the Qwen2.5 architecture, created by HauhauCS, which has undergone aggressive - [Rakazo](https://aitermbase.com/?p=89): rakazo is an open-source alternative to Grok Bot that allows users to choose their own AI model and run it in a sandboxed environment. - [Shuohao Skills](https://aitermbase.com/?p=90): eternityspring/shuohao-skills is an open-source collection of AI agent skills designed to automate the creation process for short dramas, including character breakdown, plot outlin - [Time To First Token](https://aitermbase.com/?p=91): Time to first token (TTFT) is a critical performance metric in large language model (LLM) serving that measures the latency between when a user submits a prompt and when the first - [Tokentab](https://aitermbase.com/?p=92): wzchav/tokentab is an open-source command-line interface tool that analyzes session logs from AI coding assistants like Claude Code, Codex, and Gemini CLI to calculate token usage - [Unlimited OCR](https://aitermbase.com/?p=93): Unlimited-OCR is an open-source optical character recognition model developed by Baidu that uses a causal language modeling architecture to extract text from images with theoretica - [Vibewatch](https://aitermbase.com/?p=94): Vibewatch is an open-source hardware and software project that implements a tactile stopwatch controller on an M5Stack device, designed to assist with "vibe coding" sessions by pro - [WAI Play](https://aitermbase.com/?p=96): WAI Play - AI is an open-source Python-based platform designed for automated testing and quality evaluation of AI-powered web games. - [Wan2.2 Animate 2 14B](https://aitermbase.com/?p=98): Wan2.2 Animate 2 14B is an open-source artificial intelligence model designed for video generation or animation tasks, released on the Hugging Face platform in 2026. - [Watermarks Remover](https://aitermbase.com/?p=99): guillaumemeyer/watermarks-remover is an open-source Python tool designed to remove provenance watermarks and metadata from AI-generated content across multiple vendor platforms. - [WeChat AI](https://aitermbase.com/?p=101): SMNETSTUDIO/WeChat-AI is an open-source TypeScript project that provides a framework for building and deploying AI-powered chatbots and automation tools specifically for the WeChat - [Xiaoma Durex Copywriter](https://aitermbase.com/?p=103): Xiaoma Durex Copywriter is an open-source Claude AI skill that generates marketing copy and visual concepts using Durex-style double-meaning methodology and eight copywriting formu - [杜蕾斯式文案与海报 Skill For Claude](https://aitermbase.com/?p=102): crawfordxx/xiaoma-durex-copywriter is a Claude skill that generates marketing copy and poster designs using Durex-style double-meaning methodology and eight copywriting formulas de ## Contact Corrections: admin@aitermbase.com