Stagehand

StagehandSince 2026-02

Stagehand is an AI-native browser automation SDK for TypeScript and Node.js created by Browserbase. It provides simple, high-level primitives (act, extract, and observe) that turn fragile CSS/XPath selectors into resilient natural language actions powered by frontier LLMs. By integrating directly with Playwright, Stagehand allows developers to mix standard deterministic browser code with dynamic AI-driven actions, creating test suites and automation scripts that never break when CSS classes or DOM hierarchies change.

Stagehand uses hierarchical DOM filtering to send compact accessibility trees to LLMs, keeping latency low and execution deterministic. Its extract method parses unstructured page contents directly into Zod schemas, while act executes complex multi-step user actions like selecting dates in dynamic calendar pickers. Stagehand can run locally on standard Playwright instances or scale horizontally across Browserbase's managed browser cloud with anti-bot bypass and session recording.

SmolagentsSince 2026-01

Smolagents is an ultra-lightweight, code-first Python framework created by Hugging Face for building, orchestrating, and executing autonomous AI agents in minimal lines of code. Rejecting the bloated, multi-layered abstractions of legacy agent libraries, Smolagents emphasizes 'Code Agents'—agents that express their reasoning and tool actions directly in executable Python code rather than rigid JSON string payloads. By letting LLMs write executable Python logic, Smolagents achieves vastly superior composability for data manipulation, mathematical operations, and complex loops while cutting prompt token overhead by up to 30%.

Smolagents natively supports both local open-source models (via Hugging Face Transformers and Ollama) and commercial frontier models (via LiteLLM, OpenAI, and Anthropic APIs). It includes pre-built modules for secure sandboxed Python execution, multi-modal web browsing agents with visual comprehension, and seamless sharing of agent tools directly on the Hugging Face Hub. The framework is fully modality-agnostic, supporting text, speech, vision, and custom API tools with zero vendor lock-in.