This topic covers agentic large language models (LLMs) and the platforms used to assemble, deploy, and govern autonomous workflows as of 2026-05-29. Agentic LLMs—examples include Claude Opus 4.8, Grok Build, and Copilot-style assistants—are purpose-built to plan, act through tools, and coordinate multi-step tasks. The market now splits into complementary categories: marketplaces for discovering prebuilt agents; engineering frameworks for building and testing stateful agents; AI automation and orchestration platforms for production workflows; and low-code/ no-code platforms that let non‑engineers compose agents. Key tools illustrate the landscape: LangChain provides engineering primitives and open-source frameworks (with state management and evaluation tooling) to build, debug, and deploy reliable agents; AgentGPT is a browser-based, no-code option for quickly launching goal‑driven agents; Replit combines an AI-enabled IDE and hosting for building and iterating agentic apps; GitHub Copilot embodies the Copilot-style model class by integrating code completions, chat, and autonomous developer workflows into editors and CI; Kore.ai targets enterprise deployments with multi-agent orchestration, governance, and observability; Google’s Gemini family supplies multimodal model APIs for richer tool use; Notion and Perplexity supply operational grounding—Notion for knowledge, automation, and integrations, Perplexity for web-grounded answers and citations. Trends to weigh: agent systems are growing more stateful and multimodal, marketplaces are enabling reuse and monetization, and enterprise demand is pushing stronger governance, observability, and evaluation primitives. Choosing between approaches depends on needs—rapid prototyping (AgentGPT, Replit), engineering robustness and stateful orchestration (LangChain, Kore.ai), or tight developer workflow integration (Copilot-style agents). This provides a practical framework for comparing agentic LLMs and the platforms that operationalize them.