Prototype multi-agent AI systems without writing orchestration code: AutoGen Studio is Microsoft's low-code interface over the AutoGen AgentChat framework. You compose teams of LLM-powered agents in a visual Team Builder, either by drag-and-drop from a component library or by editing the declarative JSON specification directly. Each agent gets a model, a prompt, tools (Python functions), and the team gets termination conditions and an orchestration pattern, sequential or LLM-driven. The Playground runs teams interactively with live message streaming between agents, a visual control-transition graph, tool-call and code-execution tracking, and pause/stop controls, which makes it a practical debugger for agent behavior. Finished teams export as JSON for use in any Python application via the TeamManager class, or serve as an API endpoint. Any OpenAI-compatible model endpoint works, including local servers like Ollama or vLLM. Microsoft labels it a research prototype: use it for prototyping and evaluation, and build production systems on the underlying AutoGen framework.

Sage Wiki turns a pile of unstructured documents into a fully interlinked, searchable wiki by running them through a five-pass LLM compiler pipeline. Inspired by Andrej Karpathy's vision of LLM-compiled knowledge bases, the pipeline processes source files through diff detection, summarization, concept extraction, image captioning, and cross-reference discovery, with parallel LLM calls and checkpoint/resume for vaults scaling to 100,000+ documents. The typed ontology graph stores entities and relations with BFS traversal, configurable relation types, multilingual synonyms, and a promotion/demotion lifecycle backed by grounding verification and consensus scoring. Multi-format ingestion handles Markdown, PDF, Word, Excel, PowerPoint, EPUB, email, CSV, images, and code files without manual tagging. LLM provider support spans Anthropic, OpenAI, Gemini, Ollama, and any OpenAI-compatible API, with per-pass model routing enabling cost optimization by assigning cheaper models to simpler tasks. The built-in MCP server exposes 17 tools over SSE transport for integration with Claude, Cursor, and any MCP-compatible agent, while native Obsidian vault overlay ensures existing note workflows remain undisrupted. Team deployment supports Git-synced shared wikis, centralized server access, and hub federation across multiple projects. Ships as a single Go binary with Docker Compose multi-arch images serving the web UI on port 3333. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Whirl puts every top AI model in a single conversation and remembers what you were working on last week, so you can switch models mid-thought without losing the thread. Ask for a report, a chart, or a small interactive page and it appears as a living artifact in a side panel, still editable while you keep chatting, and shareable by link when it is ready to leave the conversation. Connect your own tools over MCP, with OAuth sign-in handled for you, or install skills that teach the app new tricks your team relies on. Long-term memory carries your preferences and project context from chat to chat, and live web search with page reading grounds answers in today's pages instead of the model's training data. Sensitive threads can be locked with a password that encrypts them on your device, where the server never sees the key and only zero-retention models may answer. Around those headline pieces sits the everyday kit: incognito mode, a message queue for when you think faster than the model replies, voice input, image generation, file attachments, folders, and sharing controls. An admin console manages the model catalog, integrations, and skills, and the interface itself switches between light and dark with your choice of accent color, installable as a mobile web app. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Describe the task in plain words, pick the sources and model it may use, and an agent plans the workflow before running it. That is the core loop: agents build editable visual workflows from a task description, execute them on demand or on a schedule, and keep everything traceable. Prefer building by hand and the canvas offers 17 preconfigured steps, branching, undo and redo, JSON import and export, plus eight starter templates. Feed it real work by importing PDF, DOCX, XLSX, CSV, JSON, and text files, or by connecting folders, HTTP endpoints, and approved local programs. Upload a screen recording and it identifies the actions with timestamps, turning a demonstrated process into an editable workflow you can refine. Model calls go to the endpoint you choose: Ollama, LM Studio, vLLM, or any OpenAI-compatible service, local or cloud with your own credentials. Every run can be inspected step by step, cancelled mid-flight, and exported as JSON, CSV, or Markdown reports. Observations from completed runs land in a knowledge and learnings store, where you review them and promote the useful ones into reusable skills for the next task. The Docker release bundles the application with FFmpeg and local Whisper support, and your documents stay on your own machine. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache License 2.0 licensed.

OpenLLM serves any large language model as an OpenAI-compatible API endpoint from a single CLI command, handling model download, backend selection, quantization, and port binding automatically. It supports the full spectrum of popular models including Llama 3.3, Qwen2.5, DeepSeek, Mistral, and Phi3, choosing between vLLM and PyTorch inference backends based on hardware capabilities. When vLLM is available, continuous batching with PagedAttention achieves up to 23x throughput improvement over naive serving, while GPTQ and bitsandbytes quantization reduces memory requirements for GPU-constrained deployments. The server exposes a RESTful API on port 3000 with full OpenAI client library compatibility, enabling drop-in replacement for commercial providers in any application using the standard chat completions format. A built-in web chat UI at the /chat endpoint provides immediate interactive testing without external clients. Custom model repositories allow teams to maintain private catalogs of fine-tuned models alongside the default repository that tracks the latest releases. Deployment workflows generate production-ready Docker images automatically, with Kubernetes manifest support for orchestrated scaling. Native integration with LangChain and LlamaIndex supports RAG pipelines, Transformers Agents enables tool-calling workflows, and HuggingFace Hub handles model discovery. Server-Sent Events enable real-time token streaming across all API endpoints. Backed by BentoML's production ML infrastructure. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

One endpoint, every model you run, and a government-grade paper trail for each request. OpenGateLLM sits in front of your AI infrastructure as the single door: applications call one OpenAI-compatible API, and the gateway decides which backend answers, a self-hosted vLLM cluster, an Ollama box in the corner, a Hugging Face TEI embeddings service, or a hosted provider when you allow one. You publish friendly model names as routers and attach real provider APIs behind them, then load-balance across those backends by shuffle or least-busy so a second GPU node starts earning its keep the moment you register it. Chat completions, embeddings, reranking, OCR, and audio transcription each get their standard endpoint, so existing OpenAI SDK code, LangChain, and LlamaIndex projects point at a new base URL and keep working. Governance is built in rather than bolted on: organizations, users, and roles control who may call which model, API keys authenticate every request, and rate limiting keeps one enthusiastic team from eating the whole cluster. Usage monitoring follows every inference, and carbon footprint estimates attach to models through their hosting zone and parameter counts. Retrieval-augmented generation against Elasticsearch or Qdrant plugs into the same surface. No chat history is stored, which is rather the point when the data is yours. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

DB-GPT lets a team talk to its data: ask a business question in plain English, and an AI agent plans the work, writes the SQL, runs it, and hands back charts and a written report. Connect MySQL, PostgreSQL, SQL Server, ClickHouse, DuckDB, Doris, StarRocks, or MaxCompute, or drop in CSV and Excel files, including several at once. Every step appears in a live task view with its query, result table, code, and console output, and the agent pauses to ask clarifying questions, with answer options, when the data cannot settle the matter. Reusable skills package proven procedures, including CSV analysis, SQL analysis, and financial report analysis, and a skill creator helps write new ones. Code executes in a sandbox, parallel sub-agents split larger jobs, and an agentic knowledge-base search answers questions over uploaded documents. Scheduled tasks run saved conversations on a cron timetable with on/off switches and execution history, while MCP connectors link tools like Notion, Linear, Tavily, and GitHub. Pick OpenAI, DeepSeek, Qwen, SiliconFlow, Ollama, or vLLM for the model, then use the application and AWEL workflow tools to build data apps of your own. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

New API is a Go-powered LLM gateway that funnels over 40 AI providers, from OpenAI and Anthropic to Google Gemini, Azure, AWS Bedrock, DeepSeek, and Mistral, through a single OpenAI-compatible endpoint with intelligent routing and enterprise cost controls. Bidirectional format conversion translates between OpenAI Chat Completions, Claude Messages, and Gemini GenerateContent APIs transparently, so clients call any provider using their preferred format while the gateway handles the rest. Routing logic supports weighted random channel selection, priority-based failover, and automatic retry on provider errors to maximize uptime. The React admin dashboard shows usage charts, token consumption tracking, per-model cost breakdowns, and real-time request monitoring. Multi-tenant architecture includes three role levels, token-based authentication, per-user quota management, and a three-phase billing system with tiered pricing via a custom expression language. Multimodal support covers text completion, vision, embeddings, text-to-speech, speech-to-text, image generation, and creative task providers like Midjourney-Proxy and Suno-API. The interface ships in five languages including English, Chinese, Japanese, and French. Docker deployment runs with SQLite or MySQL for persistence. 45,000+ GitHub stars. AGPL-3.0 licensed.

To pull a team's skills, MCP servers, hooks, prompts, and agent files out of scattered Git repositories, Observal keeps them in one reviewed registry and traces real sessions to show which ones help. Build an agent visually, with a live config preview for every harness. Submissions wait in an admin review queue, where reviewers inspect side-by-side version diffs before approving a release. Teamspaces give each team its own namespace, and approved releases can be forked into independent drafts. One pull command then writes the right config for Claude Code, Cursor, Kiro, Codex, Copilot, Pi, OpenCode, Antigravity, or Goose. Every approved resource is discoverable through the open Agentic Resource Discovery standard, and a running agent can hand part of its task to another approved agent or a remote Agent2Agent service. Hooks record each session as a trace of prompts, tool calls, MCP round trips, tokens, and errors, with turn-by-turn replay for debugging and audits. Insight reports, generated through providers such as Anthropic, OpenAI, or Ollama, list what is working, what is hindering, and quick wins. A dashboard follows adoption, cost, and department usage, a leaderboard ranks agents by downloads, and audit logs, SAML single sign-on, and SCIM provisioning are included. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

With over 82,000 GitHub stars and 700,000+ downloads, LobeHub has evolved from its origins as LobeChat into a comprehensive multi-agent AI collaboration platform where humans and autonomous agent teams co-evolve. The platform's Agent Harness architecture functions as an operating system between AI models and applications, handling prompt presets, tool orchestration, lifecycle hooks, planning, filesystem access, and sub-agent management across 25+ model providers including OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Mistral, Groq, AWS Bedrock, Azure OpenAI, and local models through Ollama. Agent Groups enable sophisticated collaboration with sequential, parallel, iterative, and debate orchestration modes, allowing multiple specialized agents to tackle complex workflows simultaneously. The Agent Builder creates production-ready agents from natural language descriptions with auto-configuration, drawing from a marketplace of 505+ pre-built agents and 10,000+ MCP-compatible skills and plugins. Pages provide collaborative document editing with multi-agent co-authoring, while Schedules automate agent runs around the clock without human supervision. The knowledge base leverages PostgreSQL with pgvector for RAG-powered retrieval, and Personal Memory gives agents transparent, editable context that evolves through continual learning. The full self-hosted stack deploys via Docker Compose with PostgreSQL, Redis, RustFS for S3-compatible storage, and SearXNG for private web search, all configurable through environment variables. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. LobeHub Community licensed.

Getting a language model into the chat apps people already use takes a single self-hosted server running AstrBot, which answers in QQ, Telegram, Discord, Slack, Lark, DingTalk, LINE, and WeCom. The model itself is a setting. OpenAI-compatible endpoints, Anthropic, Google Gemini, DeepSeek, Ollama, and LM Studio all work, with fallback models and retry limits for failed requests, and Dify, Coze, or Alibaba Cloud Bailian can take over a conversation too. A bare model cannot act, so the built-in agent calls MCP servers and Skills, delegates to SubAgents, runs scheduled tasks, and executes code in an isolated sandbox. It cannot read your files either, so a knowledge base takes PDF, Word, Markdown, and EPUB uploads and searches them with hybrid dense and BM25 retrieval. Personas shape each bot's personality, and Whisper or Edge TTS adds speech. When an answer looks wrong, the ChatUI shows thinking, tool calls, token counts, and latency, and a per-message trace timeline shows how the bot got there. Because the panel steers the bot, administrator IDs, session whitelists, optional TOTP login, and HTTPS protect it. Backups live under System Settings, and the plugin market installs any of more than 1,000 extensions in one click. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

Backed by Y Combinator and powering scraping infrastructure across 195 countries, Anakin delivers a production-grade web scraping API purpose-built for AI agents and RAG pipelines that need clean, structured data from sites that actively block conventional scrapers. The single Go binary server handles JavaScript-heavy SPAs through its Camoufox anti-detect browser service with automatic fingerprint rotation, while the HTTP-first handler chain tries lightweight extraction before escalating to full browser rendering — keeping response times under 2 seconds for static pages. The built-in React 19 dashboard provides visual scraping with live results, job tracking with status filters, domain configuration management with handler chain CRUD, and proxy performance monitoring via Thompson Sampling scoring. Structured JSON extraction leverages Gemini AI to transform raw HTML into typed schemas without manual selector maintenance. SDKs span Python, TypeScript, Go, .NET, Java, and Ruby, while the MCP server exposes all 21 tools directly to Claude, Cursor, Windsurf, and any Model Context Protocol-compatible agent. The hosted platform extends the open-source engine with AI web search returning full page content with citations, multi-source agentic research across 20+ sources per query, Wire pre-built actions covering 944 websites with 5,201 structured endpoints, persistent browser sessions for authenticated scraping, and website change monitoring with scheduled alerts. Deploy via Docker Compose with three containers or run the binary directly with optional PostgreSQL persistence. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. AGPL-3.0 licensed.

Kotaemon is a document QA platform that combines advanced RAG techniques with a clean Gradio-based web interface for chatting with your documents. Built by Cinnamon, the Python backend supports any LLM provider including OpenAI, Azure OpenAI, Cohere, Groq, and local models via Ollama and llama-cpp-python, with a model management panel for configuring LLM and embedding providers from the UI. The default hybrid RAG pipeline combines full-text keyword retrieval with vector similarity search and applies re-ranking to ensure optimal result quality, while multi-modal document parsing extracts content from tables and figures alongside text. Advanced citations link every answer to specific source passages with relevance scores, viewable directly in the built-in PDF viewer with highlighted text spans. GraphRAG indexing via NanoGraphRAG, LightRAG, or Microsoft GraphRAG builds knowledge graphs from document collections for relationship-aware retrieval. Agent-based reasoning supports question decomposition for multi-hop queries using ReAct and ReWOO strategies. Multi-user authentication organizes documents into private and public collections with sharing and collaboration features. The platform supports Docker deployment in lite, full, and Ollama-bundled variants, runs on port 7860, and stores application data in a persistent volume. MCP tool integration enables external system connections for extended retrieval capabilities. On RepoCloud, deploy Kotaemon on a dedicated VPS with Docker, root SSH access, and complete control over your document AI infrastructure, all under the Apache 2.0 license.

Kortix Suna is an AI management system where autonomous agents run on isolated Linux sandbox computers, producing finished deliverables that humans review through a change request workflow before anything merges. With 20,000+ stars, it positions itself against Claude Cowork and ChatGPT Work by storing every agent persona, skill, memory artifact, and connector in a git repository: versioned, diffable, and shared across an organization. Each session launches a dedicated sandbox with full terminal access, Playwright-controlled Chromium, writable filesystem, and internet connectivity. Over 3,000 app connectors are available through MCP, OpenAPI, GraphQL, and raw HTTP, with credentials brokered server-side so tokens never enter the sandbox. Skills (reusable markdown-plus-script packages encoding company workflows) load automatically into every session, compounding institutional knowledge over time. Bring-your-own-key model routing through LiteLLM connects to OpenAI, Anthropic, Google, Mistral, or local models without vendor lock-in. The deployment runs as a single Docker Compose stack bundling the Next.js frontend, FastAPI backend, Supabase, Redis, and Caddy with automatic TLS certificates. Enterprise features include SAML/OIDC SSO, SCIM provisioning, RBAC, and audit logging. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Elastic License 2.0.

An agent run can finish without errors and still give the wrong answer, so TraceRoot traces every model and tool call, flags the production runs that missed the mark, and helps your coding agent ship verified fixes. Each run becomes a browsable call tree with inputs, outputs, latency, and cost, searchable by session or filter. Instrument with the Python or TypeScript SDK, the agent-native CLI, or any OpenTelemetry exporter, with ready integrations for LangChain, LangGraph, OpenAI Agents SDK, CrewAI, Vercel AI SDK, Mastra, and Pydantic AI. Detectors use LLM judges to score live traffic against the behavior you define and can post alerts to Slack. Judge outputs cluster into signals, then a background coding agent reads the traces and your connected GitHub code in a sandbox and recommends changes down to the commit, file, and line. Datasets, test cases, and scored eval runs compare versions so regressions surface before a release, and dashboards of widgets chart your trace data. The CLI hands your coding agent, such as Claude Code or Codex, the traces, signals, detectors, and evals to work from. Self-hosting keeps all of that trace data on infrastructure you control. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed (core edition).

Casibase lets organizations build AI-powered knowledge bases that answer questions from their own documents, connecting to 30+ model providers through a unified admin interface with RAG retrieval and multi-agent orchestration via MCP and A2A protocols. The platform plugs into OpenAI GPT-4o, Anthropic Claude, Meta Llama, Google Gemini, DeepSeek, Ollama local models, HuggingFace, Azure OpenAI, and additional providers, while embedding APIs from OpenAI Ada and Baidu handle vector representation of ingested documents. Document ingestion parses TXT, Markdown, DOCX, PDF, CSV, XLSX, and PPTX files with intelligent chunking strategies for optimal retrieval accuracy. The built-in chat interface provides real-time AI conversations with manual session handover for human agent escalation, and comprehensive chat session logging enables audit trails for compliance. Enterprise identity management integrates Casdoor for Single Sign-On supporting GitHub, Google, WeChat, and OIDC providers with fine-grained access control via the Casbin permission engine. The multi-tenant architecture supports isolated knowledge bases per organization with role-based user management and configurable storage, model, and embedding providers per tenant. The React frontend with Ant Design v5 provides a polished admin dashboard for managing providers, knowledge stores, chat sessions, and user access, while the Go backend with Beego framework handles API logic with MySQL or MariaDB persistence. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

Accepted as a NeurIPS 2025 Spotlight paper and rapidly approaching 6,000 GitHub stars, AI-Researcher from the Hong Kong University Data Science Lab delivers the first fully autonomous scientific research system — a multi-agent platform that takes a list of reference papers and returns a complete research contribution with working code, validated experiments, and a formatted academic manuscript. The pipeline orchestrates five distinct phases: a Resource Collector systematically gathers materials from arXiv, IEEE Xplore, ACM Digital Library, Google Scholar, GitHub, and Hugging Face; an Idea Generator performs gap analysis against semantic embeddings to produce 3-5 novel hypotheses with feasibility scores; an Algorithm Designer transforms concepts into functional implementations; a Validation Engine automates testing, performance evaluation, and iterative optimization; and a Manuscript Creator generates polished full-length papers with figures, tables, and citations. The Gradio-based web GUI provides intuitive tabs for environment configuration, example selection, and real-time monitoring of research progress, while the production deployment at novix.science offers immediate browser access without local setup. Scientist-Bench provides a standardized benchmark comprising state-of-the-art papers across diverse AI research domains for evaluating autonomous research capabilities. The system supports multiple LLM providers including OpenAI, Anthropic, Google Gemini, and OpenRouter models with per-task routing for cost optimization. Deploy via Python with pip dependencies or Docker containerization. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Anyone who pays for Claude Code, Codex, and Gemini CLI separately ends up juggling three endpoints, three sets of keys, and three kinds of rate-limit surprise, and CliRelay collapses all of that into one self-hosted gateway with a control panel. Clients speak the OpenAI Chat or Responses, Anthropic Messages, or Gemini protocol, and CliRelay translates each request for whichever upstream account is healthy. Channel groups limit where each key may send traffic, round-robin, fill-first, and session-sticky scheduling spread it, and cooldowns with automatic failover move requests off accounts that have run out of quota. Add upstreams from the /manage console through browser sign-in, device codes, or bulk credential import, covering Claude, Codex, Gemini, Antigravity, Vertex AI, AWS Bedrock, Grok, Qwen, Kimi, and any OpenAI-compatible service. Portal accounts own client keys with daily, period, and lifetime quotas, RPM and TPM limits, and per-model pricing. A monitor center scores health and charts live traffic, P50 to P99 latency, and time to first token, while every request lands in a searchable log with tokens, latency, and cost. Tenants, roles with resource.action permissions, content moderation profiles, an outbound proxy pool, and an audit trail make shared deployments workable, and the console speaks English, Simplified Chinese, and Russian. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.