Google AI Edge Gallery

Google AI Edge Gallery brings on-device Generative AI to your phone-run models offline, switch Hugging Face models, ask about images, transcribe/translate audio, test prompts and view real-time performance metrics.

ROMA

ROMA is an open-source framework for creating customizable multi-agent AI systems. It supports any model and data source, enabling development of versatile agents like task solvers, research assistants, and finance advisors for specialized applica...

Cognitora

Cognitora provides secure, scalable compute for autonomous AI agents: persistent code execution plus containerized microVMs (Kata on Firecracker), sub-second cold starts and LangChain/AutoGPT integrations.

Grok Code Fast 1

Grok Code Fast 1 is a fast, cost-effective AI coding model integrated into top platforms like GitHub Copilot and Cursor, offering reliable daily support for agentic coding with an available Prompt Engineering Guide for optimized results.

GitArsenal

GitArsenal uses AI to diagnose and fix dependency and hardware issues in existing codebases, ensuring your projects run smoothly without hours of manual debugging. It automatically provisions the right environment and keeps troubleshooting until y...

Gitmore

Gitmore syncs with GitHub and Bitbucket to deliver AI-powered daily or weekly repo reports via Slack or email, helping teams track PRs, follow-ups, and updates effortlessly without sifting through dashboards or manual data.

TensorZero

TensorZero is an open-source LLM infrastructure offering a unified API for multiple providers, real-time observability, optimization through feedback, evaluation tools, and built-in experimentation to streamline industrial-grade large language mod...

Hierarchical Reasoning Model

Hierarchical Reasoning Model solves complex problems like extreme Sudoku and large-scale pathfinding using just 1,000 examples without pre-training. It outperforms larger models on reasoning tasks with fewer parameters and minimal data.

BashNota

BashNota is an open-source tool that simplifies managing and organizing Bash scripts, enabling clear documentation and efficient code maintenance for developers working with shell scripting.

GLM-4.5

GLM-4.5 is a 355B parameter open-weight MoE model delivering advanced reasoning, coding, and agentic task performance. Available in flagship 355B and 106B Air versions, it supports efficient dual-mode inference for versatile AI applications.

Grok CLI (Unofficial)

Grok CLI (Unofficial) lets you control your terminal using natural language commands. It selects the right tools automatically and supports custom project instructions, enabling faster workflows without memorizing complex syntax.

Palmier

Palmier is an autonomous AI assistant for software engineering that automates tasks triggered by GitHub or Slack events. It writes code, fixes bugs, reviews PRs, and triages issues, helping teams save time and ship software faster with full code c...

VoltOps

VoltOps provides framework-agnostic observability for AI agents and LLM apps, enabling teams to trace, debug, and monitor every agent step with live logs and clear dashboards for improved transparency and performance.

GitProbe

GitProbe transforms any GitHub repo into a clear summary of its architecture by mapping function-level relationships. Get a breakdown of core modules, key functions, and data flow in plain text to speed up codebase onboarding and analysis.

systemprompt MCP

systemprompt MCP is the first voice-controlled app for managing multiple Model Context Protocol servers on iOS and Android, enabling developers and DevOps to control AI agents, monitor errors, and review pull requests hands-free, anytime and anywh...

VibeSec

VibeSec is an AI-powered code security copilot that scans GitHub repositories instantly, detects vulnerabilities, and delivers clear reports to help developers and teams prevent security issues before deployment.

MiniCPM 4.0

MiniCPM 4.0 is an open-source family of ultra-efficient on-device AI models offering fast performance on edge chips. It features highly quantized BitCPM versions for enhanced speed and efficiency in resource-constrained environments.

Duelin' Agents: AI Chat Face-Off

Duelin' Agents enables two AI models to chat live via split-screen. Configure GPT-4, Claude, and more with webhooks and n8n support to set prompts and watch agents debate, create, and collaborate in real time with turn-based controls.

Athena

Athena by Microsoft is an AI agent in Teams that streamlines developer workflows by managing pull requests, tasks, and releases. It boosts collaboration and efficiency with an open-source template (Dex) for easy customization.

GitHub Chat

GitHub Chat lets you instantly query any GitHub repository, file, or wiki by replacing "github.com" with "githubchat.ai," providing quick, AI-powered answers about project documentation, code, or wiki content for efficient development.

Circuit Tracer

Circuit Tracer by Anthropic visualizes internal computations of large language models as attribution graphs, enabling researchers to analyze and interpret AI decision processes for greater transparency and insight. Available on Neuronpedia and as ...

HunyuanVideo-Avatar

HunyuanVideo-Avatar by Tencent generates emotion-controlled, multi-character talking avatar videos from audio. This open-source tool ensures consistent character representation with released code and models for seamless video creation.

Jazzberry

Jazzberry is an AI agent that detects bugs by executing real code. Integrated with GitHub, it automatically generates bug reports on your pull requests, helping you catch issues early and improve code quality efficiently.

Visio Driver Assistant

Visio Driver Assistant detects objects on the road and provides timely alerts, from gentle nudges to clear warnings, adapting intelligently to your driving conditions to help keep you safe and aware.

Bagel

Bagel by ByteDance-Seed is an open-source multimodal AI model that enables advanced image and text generation, editing, and interaction, offering capabilities on par with proprietary solutions under the Apache 2.0 license.

Magentic-UI

Magentic-UI by Microsoft Research is an open-source web agent that enables collaborative planning and execution of web tasks with AI. It features co-planning, action guards, and plan learning to streamline complex workflows.

Jules

Jules by Google Labs is an asynchronous AI coding agent for GitHub repos that analyzes project context to fix bugs, build features, and generate audio changelogs. Public beta available now.

GitHub Copilot Coding Agent

GitHub Copilot Coding Agent automates coding tasks by implementing issues and running in the background with GitHub Actions, streamlining development and boosting productivity with seamless AI-assisted workflows.

Ollama v0.7

Ollama v0.7 offers a new engine for first-class multimodal AI, integrating vision models like Llama 4 and Gemma 3. It enhances reliability, accuracy, and memory management for efficient local LLM deployment.

Codex by ChatGPT

Codex by ChatGPT is a cloud-based agent that writes features, fixes bugs, answers code questions, and suggests pull requests—all within isolated sandbox environments preloaded with your repository for efficient development.

VibeShift MCP

VibeShift MCP uses an auto feedback loop to generate secure, fully functional code in a single attempt. This open-source tool streamlines coding by minimizing errors and boosting efficiency in software development.

HanaVerse

HanaVerse is an open-source, self-hosted web UI that brings Ollama language models to life through Hana, a 2D anime character who visually reacts and provides spoken responses via text-to-speech for interactive conversations.

Tersa

Tersa is an open source canvas for building AI workflows by dragging, dropping, and connecting nodes. Easily create custom workflows powered by top AI models to streamline and automate complex tasks.

AG-UI

AG-UI is an open, lightweight protocol that standardizes AI agent connections to front-end apps, enabling real-time interactivity, streaming, and human-in-the-loop workflows with support for LangGraph, CrewAI, and more.

Line0

Line0 is an AI pair programmer for backend developers, enabling fast creation of fully functional Express.js API services. It features seamless GitHub integration, an in-browser code editor, and enhanced prompts to streamline backend development.

OpenMemory MCP

OpenMemory MCP is a private, local-first memory layer with a built-in UI that keeps all data structured and fully controlled on your device, ensuring no cloud sync or external storage for maximum privacy and security. Compatible with all MCP-clients.

Phala Cloud

Phala Cloud offers a zero-trust confidential compute platform for AI teams to build and run safe AGI. Launch TEE CPU and GPU instances in seconds with end-to-end AI encryption. Enjoy 30 days of free credits on launch.

GitBrain

GitBrain streamlines your Git workflow by using AI to analyze code changes and suggest the best Git operations, helping you save time and boost productivity in version control management.

RM Flags

RM Flags automates feature flag cleanup by detecting stale flags in your codebase. It integrates with version control, sends notifications, and creates automated PRs, ensuring your code stays clean and manageable with professional support.

CodebaseChat

CodebaseChat lets you interact with any GitHub repository by creating an OpenAI GPT trained on its code. Simply provide the repo URL to ask questions and get insights. Fully free and open source for seamless code exploration.

LangChain Templates

LangChain Templates lets you find, download, and customize templates for various GenAI models and use cases, all managed conveniently in one platform to streamline your AI development process.

Flow by Laminar [LW24]

Flow by Laminar [LW24] is a lightweight task engine that simplifies building AI agents with clean code, enabling easy creation of parallel, stateful, and dynamic workflows for efficient and scalable agentic applications.

Cloudy

Cloudy auto-generates and updates technical documentation instantly with every code change, ensuring your docs stay accurate and aligned with your latest pull requests. Streamline documentation without extra effort.

Rememberall

Rememberall is a secure, open-source tool that enables Custom GPTs to retain memory across conversations. Mention @rememberall to access past context and maintain seamless, continuous interactions with your AI.

wrapped.dev

wrapped.dev uses AI to create detailed summaries of GitHub repositories, highlighting developers' open-source contributions and the effort behind projects, making it easier to appreciate and share code work.

Copy Repo

Copy Repo lets you instantly copy an entire GitHub repository with one click, enabling seamless integration and easy access for working with your preferred large language model. Save time and streamline your coding workflow efficiently.

GitHub Spark AI

GitHub Spark AI lets you create and share customizable micro apps without coding, accessible on desktop and mobile. Streamline tasks and workflows by building sparks that fit your specific needs quickly and easily.

StarLens

StarLens analyzes your GitHub repo stars and provides AI-driven insights. Customize outputs with AI Model and Voice/Tone settings. Built with low-code tools, it’s easy to clone and create your own AI-powered workflows quickly.

WorkWeave

WorkWeave connects to your engineering workflows and uses AI to deliver personalized insights and actionable feedback, boosting productivity and turning any engineer into a 10x contributor.

MUI for VS Code

MUI for VS Code lets you chat with an AI expert on MUI directly within VS Code. Type "@mui" in GitHub Copilot to get accurate, up-to-date answers with references from the latest MUI documentation, streamlining your development process.

ChattyUI

ChattyUI is an open-source, feature-rich chat interface that runs open-source AI models like Gemma and Mistral locally in your browser via WebGPU, ensuring all processing happens on your PC with no data sent to external servers.

Weekly Github

Weekly Github provides a clear, AI-generated summary of your GitHub activity from the past week, helping you stay updated on your progress and contributions without missing key developments.

HyperCrawl

HyperCrawl is a zero-latency web crawler optimized for retrieval-based large language model development, enabling instant and efficient data extraction to support accurate and timely AI training and information retrieval.

Prompto

Prompto is an open-source web app that lets you easily chat with, query, and fine-tune multiple large language models. Save and manage prompts efficiently to streamline your AI interactions in one user-friendly platform.

Warpy: Terminal Generative Shell

Warpy: Terminal Generative Shell blends Rust’s efficiency with GPT-powered natural language to create an interactive, conversational terminal experience that simplifies command execution and enhances user productivity.

Super RAG

Super RAG combines summarization, retrieval, reranking, and code interpretation into a single API, streamlining data processing and enhancing accuracy for efficient information extraction and analysis.

AlphaGeometry from Google DeepMind

AlphaGeometry from Google DeepMind solves challenging Olympiad-level geometry problems with accuracy approaching top human gold-medalists, using AI trained solely on synthetic data to advance mathematical reasoning capabilities.

Amica

Amica is an AI tool that streamlines project management by automating task tracking, deadlines, and team collaboration, helping businesses improve efficiency and meet goals with ease.

UI-TARS Desktop

UI-TARS Desktop by Bytedance lets you control your computer through natural language commands using a vision-language model, simplifying interaction and boosting productivity with an intuitive GUI agent.

VisionAgent

VisionAgent enables precise, reasoning-driven object detection using simple text prompts—eliminating the need for custom training. Created by Andrew Ng's Landing AI, it delivers human-like accuracy efficiently and effectively.

QWQ-Max

QWQ-Max by Qwen is a versatile large language model excelling in reasoning, math, coding, and agent tasks. Its unique "thinking mode" enhances problem-solving for complex challenges. Open-source release coming soon.

Pull Sense

Pull Sense automates pull request reviews by providing AI-generated feedback to identify bugs, security risks, and code improvements. It supports Bring Your Own Key (BYOK) and complements human reviewers for safer, more efficient code reviews.

ExplainGithub

ExplainGitHub uses AI to analyze GitHub repositories, helping developers quickly grasp unfamiliar codebases and improve productivity by simplifying code exploration and comprehension.

Graphiti

Graphiti builds dynamic knowledge graphs from evolving business data and chat histories, enabling Python AI agents to quickly access accurate, relevant information for improved decision-making. Explore more on our GitHub repository.

DeepSeek-VL2

DeepSeek-VL2 offers open-source vision-language models with efficient MoE architecture, delivering strong multimodal capabilities. Test its performance easily through the Hugging Face demo for advanced image and text integration.

Instella

Instella by AMD delivers high-performance 3B language models trained on MI300X, offering efficient, open-licensed AI solutions with ResearchRAIL model weights and MIT-licensed code for advanced language processing tasks.

HunyuanVideo-I2V

HunyuanVideo-I2V by Tencent is an open-source image-to-video model that generates up to 720p, 129-frame videos. It supports custom LoRA training for creating unique visual effects, enabling flexible and high-quality video synthesis.

RLAMA

RLAMA is an AI assistant that enhances local data retrieval by combining retrieval-augmented methods with a local model, enabling efficient, accurate, and private access to relevant information without relying on external servers.

TxGemma

TxGemma, an open model based on Google's Gemma 2, specializes in therapeutics development. It predicts molecule properties, analyzes data, and facilitates result discussions. Available on Hugging Face and Vertex AI for seamless integration.

Haystack Code Reviewer

Haystack Code Reviewer organizes and analyzes pull request changes, guiding you through them on an interactive canvas. It streamlines code reviews, helping you focus on meaningful feedback and understanding updates efficiently.

Orpheus TTS

Orpheus TTS is an open-source text-to-speech tool powered by Llama-3b, delivering natural, human-like speech with emotional nuance. It offers zero-shot voice cloning, guided emotion control, and low-latency streaming for seamless audio generation.

Agent Development Kit

Agent Development Kit (ADK) is an open-source platform for building powerful multi-agent systems, offering flexible orchestration, a rich tool and model ecosystem, and built-in evaluation to streamline development and optimize performance.

Dream 7B

Dream 7B is a leading open diffusion large language model that matches or surpasses similar-sized AR models like LLaMA3 and Qwen2.5. It excels in advanced planning and offers flexible inference for versatile AI applications.

GoBuildMy.App

GoBuildMy.App lets you create native iOS and Android apps by simply chatting. Quickly turn ideas into functional apps, connect APIs with prompts, build MVPs in minutes, and automatically back up all code to GitHub for seamless development.

Meta Perception Encoder

Meta Perception Encoder is a cutting-edge vision encoder that sets new benchmarks in image and video analysis. It excels in zero-shot classification and retrieval, delivering superior accuracy and performance beyond existing models.

MiMo

MiMo is an open-source LLM series focused on reasoning, offering pre-trained and RL-tuned models that deliver strong math and coding performance comparable to o1-mini, with base, SFT, and RL variants available.

DeepWiki by Congnition

DeepWiki by Cognition uses AI to generate interactive documentation for GitHub repos, enabling natural language queries about codebase structure. Powered by Devin, it’s free for open-source projects, simplifying code understanding and collaboration.

InternVL3

InternVL3 from OpenGVLab is an advanced multimodal AI model excelling in vision, reasoning, and long-context understanding. It enhances performance across text and visual tasks, outperforming base LLMs with native multimodal pre-training.

GitDiagram

GitDiagram converts any GitHub repository into an interactive diagram in seconds, providing a clear visual overview of project structure and relationships to streamline code analysis and collaboration.

Product Hunt MCP

Product Hunt MCP integrates Product Hunt's API with LLMs via the Model Context Protocol, enabling seamless access to posts, collections, topics, users, votes, and comments through Claude Desktop, Cursor, or any MCP-compatible client.

GitHub MCP Server

GitHub MCP Server enables AI agents to securely access GitHub APIs locally, streamlining development workflows. Seamlessly integrates with VS Code for efficient, secure API calls within your coding environment.

Potpie AI

Potpie AI builds custom task-oriented agents that leverage your codebase and data context to perform precise engineering tasks like system design, debugging, testing, and onboarding, enhancing development efficiency and accuracy.

Portia AI

Portia AI enables fast deployment of secure, production-ready AI agents with plug-and-play integrations, declarative planning, and unified authentication for full human oversight across API and web platforms.

Agno

Agno is a lightweight library for building advanced Reasoning Agents with integrated memory, knowledge management, and multi-modal support, enabling smarter, context-aware AI applications with seamless tool integration.

GitSummarize

GitSummarize transforms any GitHub repository into an AI-powered, always up-to-date documentation hub. It eliminates outdated docs, ensuring accurate, interactive summaries instantly—just replace 'hub' with 'summarize' in the repo URL.

Qwen2.5-Omni

Qwen2.5-Omni by Alibaba Cloud is a multimodal AI model that processes text, images, audio, and video, enabling seamless generation of text and natural streaming speech for versatile communication and content creation.

Skywork-R1V

Skywork-R1V is an open-source multimodal reasoning model specialized in visual math, science, and complex problem-solving, delivering accurate and efficient analysis across diverse scientific and mathematical tasks.

GitHub Copilot for Raycast

GitHub Copilot for Raycast lets you start and track GitHub Copilot coding agent tasks directly from Raycast, streamlining your workflow and improving task management without leaving your command interface.

Augment Agent

Augment Agent is an AI pair programmer that deeply understands your codebase with 200K context and advanced memory. Open-sourced with #1 SWE-bench approach, it seamlessly integrates into VS Code and JetBrains for smarter, faster coding.

Solver

Solver is an autonomous coding agent that completes software tasks independently. It works directly in your git repositories, handling bug fixes and new features so you can assign work and return to ready-to-review code.

ApertureDB Multimodal AI Workflows

ApertureDB Multimodal AI Workflows streamlines generating embeddings, object detection, attribute inference, and querying across multimodal datasets. Ingest, enrich, and analyze your data efficiently using integrated Jupyter notebooks.

Animate Anyone 2

Animate Anyone 2 is an open-source AI model by Alibaba’s Tongyi Lab that delivers high-fidelity character animations with realistic interactions in dynamic environments for enhanced digital storytelling and simulation.

BashBuddy

BashBuddy generates bash and PowerShell commands from natural language input, boosting productivity. It runs locally or via an optional $2 cloud service for instant results. Fully open-source for transparent, efficient command creation.

MARS5 TTS

MARS5 TTS is an open-source text-to-speech model that replicates voice performances from just 2-3 seconds of audio across 140+ languages, handling complex prosody like sports commentary, movies, and anime with high accuracy. Join our Discord to le...

Langflow Desktop

Langflow Desktop is a developer tool that enables visual building and deployment of AI-powered agents. It supports major LLMs, vector databases, and offers an API server to seamlessly turn agents into accessible APIs.

Build Y

Build Y delivers the latest open-source engineering and technology advancements, streamlining access to innovative tools and resources for efficient project development and innovation.

a0.dev Phase 1

a0.dev Phase 1 delivers the fastest iOS builds with one-click App Store submission. Featuring AI agent thinking mode, Stripe payments, and GitHub sync, it transforms your ideas into apps quickly and efficiently.

OmniParser V2

OmniParser V2 converts UI screenshots into structured, tokenized elements that large language models can interpret, enabling precise next-action predictions based on interactable components extracted directly from pixel data.

Unsloth

Unsloth enables 2x faster fine-tuning of LLMs like Llama 3 and Mistral with up to 80% less memory usage. Open-source and equipped with reasoning capabilities, it offers free Colab notebooks for efficient model customization.

Promptless

Promptless is an AI teammate that keeps your documentation up to date by automatically drafting updates when new Pull Requests, resolved support tickets, or Slack tags occur, enhancing customer experience and minimizing lost revenue.