MIT agent runtime. Pay model API only. Live Flash peak $0.44/$1.32 per 1M, not the old $0.14/$0.28 row.

Open multi-agent framework. CLI is $0 + BYOK. MGX is now Atoms: Free $0 / Pro from $20.

Ultra-lightweight, self-hosted personal AI agent framework in Python with WebUI, tools, long-term memory, MCP integrations, multi-agent workflows, and scheduled automation.

MIT agent framework. Hosted traces are LangSmith Developer $0 / Plus $39 per seat. Not LangServe-as-the-product.

Open-source AI agent framework developed by Luki Labs, providing flexible agent orchestration and workflow management for building complex AI applications with multi-agent collaboration.

Vercel's open framework for building production AI agents. Like Next.js for agents: instructions and skills in Markdown, tools in TypeScript, durable by default.

Enhanced fork of Craft Agents with GUI for Skills management, session import/export, one-click installation, no manual config editing required. Optimized for Claude Code + Agents workflows, making AI Agents easy for anyone to use.

Meland Labs workspace that turns expert methods into a self-evolving digital employee. Early access v0.6.1. Free, Basic $15/mo, Pro $39/mo.

Anthropic's official plugin pack for finance: reference agents and skills for pitch decks, DCF/LBO models, earnings review, and GL reconciliation.

Self-improving open-source AI agent by Nous Research with a built-in learning loop, multi-channel gateway, and model-agnostic runtime. MIT licensed.

Open-source TypeScript agent framework: agents, workflows, RAG, memory, MCP, and eval with a hosted platform. About 27,437 GitHub stars (2026-08-25).

NVIDIA's open-source agent framework that turns an agent into one Python class: methods are actions, docstrings are prompts, at 82.2% SWE-bench Verified.

Vercel Labs' MIT-licensed reference app for running background coding agents: a web UI, a durable workflow runtime, and isolated sandboxes you can fork.

AWS's batteries-included agent harness: one import gives you a tuned general-purpose agent that the team says runs 28 percent cheaper than Claude Code.

Self-evolving heavy-duty solver for deep research and agentic work. Splits reasoner from verifier, runs a trained Agent Team, and ships open weights plus the FrontierAgent CLI research harness.