AgDex.ai — AI Agent Tools Directory
A curated directory of 708+ AI agent tools, frameworks, LLM APIs, cloud infrastructure, and developer utilities. Updated regularly to cover the latest in the AI agent ecosystem.
Featured AI Agent Tools
Developer Tools & Observability
- Claude Code — Anthropic's agentic coding tool operating in your terminal. 80.9% SWE-bench score. Reads codebase, writes/runs code, and deploys with full git integration.
- LangSmith — Official LangChain observability platform for tracing, debugging and evaluating LLM apps. Deep LangChain/LangGraph integration.
- Weights & Biases — ML experiment tracking and model management. Supports hyperparameter tuning, dataset versioning, and LLM fine-tuning.
- Streamlit — Build data apps and AI interfaces in pure Python. No frontend skills needed for interactive web apps. By Snowflake.
- MLflow — Open-source ML lifecycle platform for experiment tracking, model registry, deployment, and LLM evaluation. By Databricks.
- Gradio — Python library for quickly building ML/AI demo UIs. A few lines of code for web interfaces; powers Hugging Face Spaces.
- Ragas — Evaluation framework for RAG pipelines. Automatically measures retrieval accuracy, faithfulness, and answer relevance.
- Chainlit — Python framework for rapidly building LLM app UIs. Few lines of code for chat interfaces with streaming and step tracing.
- Aider — AI pair programming in your terminal — edit code with GPT-4o, Claude, and other LLMs directly in git repos
- Augment Code — AI coding agent for large codebases (400K+ files). 70.6% SWE-bench. Deep codebase indexing, VS Code & JetBrains extensions, autonomous agent mode.
- Amazon Q Developer — AWS AI coding assistant. Inline completions, autonomous code transformation, security scanning, CLI help, and deep AWS integration. Free and Pro ($19/user/mo) tiers.
- vLLM — Fast LLM serving with PagedAttention. Continuous batching, high throughput, OpenAI-compatible API. The production standard for self-hosted LLM inference. 30K+ stars.
- Elasticsearch — Distributed search and analytics. Full-text + vector (HNSW) + semantic search in one engine. Enterprise RAG backbone with 8.x ELSER sparse vector model.
- DeepEval — Unit testing framework for LLM apps with 14+ built-in metrics. Hallucination detection, RAG evaluation, works like Pytest.
Ecosystem & Orchestration
- LlamaIndex — Data framework connecting LLMs to external data. RAG pipelines, data connectors, indexes, and query engines for knowledge apps.
- Midjourney — Leading AI image generation model. Artistic quality and aesthetic coherence. V7 supports editing, character refs, and personalized style. Web UI + Discord.
- Stable Diffusion — Stability AI's open-source image generation. SD3.5/SDXL with vast fine-tune/LoRA ecosystem. ComfyUI and AUTOMATIC1111 WebUI for local use.
- Dify — Open-source LLM app platform. Visual workflow for RAG, agents, chatbots. 60K+ GitHub stars. Self-host or Cloud. 50+ model providers.
- Semantic Kernel — Microsoft open-source AI orchestration SDK for C#/Python/Java. Seamlessly integrates AI models with existing code.
- FLUX — State-of-the-art image generation by Black Forest Labs. FLUX.1 Pro/Dev/Schnell lead quality benchmarks. FLUX.1 Kontext for in-context image editing.
- Microsoft 365 Copilot — AI in Word, Excel, PowerPoint, Teams, and Outlook. Drafts docs, analyzes data, summarizes meetings, and answers questions over company knowledge.
- Haystack — Open-source LLM app framework by deepset. Pipeline abstraction, modular RAG, and agent building for production use.
- HeyGen — AI video platform with photorealistic avatars. Create videos from text, translate with lip-sync in 40+ languages, build interactive AI twins.
- PydanticAI — Agent framework by the Pydantic team. Type-safe, native dependency injection, seamless Pydantic ecosystem integration.
- Google Workspace AI — Gemini AI in Gmail, Docs, Sheets, Slides, and Meet. Summarizes emails, drafts content, analyzes data, creates presentations, and transcribes meetings.
- Botpress — Enterprise AI chatbot building platform with visual conversation flow design, multi-channel deployment, and GPT integration.
- Flowise — Open-source drag-and-drop UI for building LLM flows and AI agent pipelines visually
- Sora — OpenAI's text-to-video model. Realistic videos up to 60s from text or images. Available in ChatGPT Plus and Pro.
- Ideogram — AI image generation with superior text accuracy. Ideogram 3.0 renders text in images (logos, posters) with high precision. Realistic + design styles, free tier.
- Synthesia — Enterprise AI video platform. 230+ AI avatars for training and marketing videos without cameras. Used by 50K+ companies for L&D and internal communications.
- Activepieces — Open-source Zapier alternative with self-hosting support, 200+ integrations, and AI features for privacy-conscious teams.
- Voiceflow — Team-oriented AI agent building platform with drag-and-drop conversation design and multi-channel deployment for customer service.
Core Frameworks
- OpenAI Agents SDK — OpenAI's official Python SDK for building multi-agent systems with handoffs, tools, and guardrails
- LangGraph — Graph-based framework for building stateful, multi-actor LLM applications and agents by LangChain
- Google ADK — Google's Agent Development Kit — open-source framework for building, testing and deploying AI agents
- OpenHands — Open-source AI software engineer agent that can write code, run commands, and browse the web autonomously
- SWE-agent — Open-source software engineering agent by Princeton. Autonomously fixes GitHub issues, top performance on SWE-bench.
- NotebookLM — Google's AI research notebook. Ground responses in your docs (PDFs, slides, URLs). AI podcasts (Audio Overview), source-grounded Q&A, and study guides.
- STORM — Stanford open-source agent for automated research reports. Multi-perspective web research generates cited Wikipedia-style articles.
- MetaGPT — Multi-agent framework assigning different roles to GPTs to collaboratively solve complex software tasks
Cloud Infrastructure
- Railway — Developer-friendly cloud platform for deploying AI agents and backends with zero-config infrastructure
- Modal — Serverless cloud platform for running Python functions on GPUs. Auto-scales, pay-per-use, ideal for AI/ML workloads.
- RunPod — GPU cloud for AI workloads — cost-effective compute for training and running AI agents at scale
- Zilliz Cloud — Fully managed vector database cloud by the Milvus team. Supports billion-scale vector search with enterprise-grade SLA.
- E2B — Cloud sandbox for AI agents — secure code execution environments for running agent-generated code
- Amazon Bedrock Agents — Fully managed AWS agent service. Build multi-step agents using Foundation Models with native integration to Lambda, S3, RDS, and 40+ AWS action groups.
- Azure AI Foundry — Microsoft's unified AI platform. 1800+ models (OpenAI, Meta, Mistral), fine-tuning, agent deployment, and enterprise monitoring with Azure security and compliance.
- Qdrant Cloud — Managed cloud version of the high-performance Qdrant vector database. Supports hybrid search for RAG and semantic search.
- Amazon Bedrock — AWS managed service for building AI agents with foundation models from Anthropic, Meta, Mistral, and more
- Fireworks AI — Ultra-low latency LLM inference platform. FireFunction supports tool calling, ideal for latency-sensitive AI agent apps.
- Vertex AI — Google Cloud's unified ML platform with Agent Builder for deploying production AI agents
- Vertex AI Agent Builder — Google Cloud managed agent platform. Visual builder, grounding with Google Search, enterprise data connectors, multi-agent orchestration, and ADK integration.
- Lambda Labs — GPU cloud for AI/ML with on-demand H100/A100 instances, persistent storage, and deep learning optimized workstations.
- Supabase Vector — Vector storage built on PostgreSQL + pgvector, seamlessly integrated with the Supabase platform. Great for RAG apps.
- Vast.ai — Global GPU rental marketplace offering 3-5x cheaper compute than major clouds. Ideal for AI training and inference.
LLM APIs & Providers
- Claude API — Anthropic's Claude API — access Claude 3.5/3.7 Sonnet/Opus for building safe and capable AI agents
- Llama 4 — Meta's open-weight MoE model family. Scout: 10M token context on single H100. Maverick: 1M context, 400B total params. Free to use, fine-tune, and deploy.
- Gemini API — Google's multimodal LLM API. Gemini 2.5 Pro/Flash: 1M+ context, native audio/video/image, code execution, Google Search grounding via AI Studio or Vertex.
- Qwen — Alibaba's open-source LLM series (Qwen3). 235B MoE + 0.6B–32B dense models. Top open-source benchmarks, 119 languages, thinking mode, Apache 2.0.
- Gemma 3 — Google's open-weight model family for on-device and research. 1B–27B sizes. Instruction-tuned for chat and agentic tasks. Apache 2.0 licensed.
Why AgDex.ai?
AgDex.ai is an independent, editorially curated directory focused exclusively on the AI agent ecosystem. Every tool is manually reviewed for relevance, quality, and active maintenance. We cover tools from across the globe — including tools popular in Japan, Germany, South Korea, France, and beyond. All descriptions are written in English, Spanish, German, and Japanese by human editors, not machine-translated.
Whether you are building your first AI agent or architecting a production multi-agent system, AgDex.ai provides a comprehensive, up-to-date reference for the tools and frameworks you need.
Contact us: agdex.ai@gmail.com | About | Blog | Contact