# Steel Documentation > Steel is a cloud browser API for AI agents and developers. > Use Steel to launch cloud browsers, scrape content, and automate web tasks. ## Quick Reference - Install: `npm install steel-sdk` (Node.js) or `pip install steel-sdk` (Python) - CLI: `curl -sSf https://setup.steel.dev | sh` - Auth header: `steel-api-key: ` - Auth env var: `STEEL_API_KEY` - API base URL: `https://api.steel.dev` - WebSocket: `wss://connect.steel.dev?apiKey=&sessionId=` - API reference: https://steel.apidocumentation.com/api-reference ## Agent Instructions - For the simplest one-liner scrape, use the CLI: ```bash steel scrape https://example.com ``` - For simple scraping without a browser session, use the REST scrape endpoint: ``` curl -X POST https://api.steel.dev/v1/scrape \ -H "steel-api-key: YOUR_KEY" \ -H "Content-Type: application/json" \ -d '{"url": "https://example.com"}' ``` - For browser automation, connect Puppeteer or Playwright via WebSocket: ```js import Steel from 'steel-sdk'; import puppeteer from 'puppeteer-core'; const client = new Steel({ steelAPIKey: process.env.STEEL_API_KEY }); const session = await client.sessions.create(); const browser = await puppeteer.connect({ browserWSEndpoint: `wss://connect.steel.dev?apiKey=${process.env.STEEL_API_KEY}&sessionId=${session.id}`, }); // ... use browser ... await browser.close(); await client.sessions.release(session.id); ``` - Python SDK: ```python from steel import Steel client = Steel(steel_api_key="YOUR_KEY") # or set STEEL_API_KEY env var result = client.scrape(url="https://example.com") print(result.content.html) ``` - `client.scrape()` returns a `ScrapeResponse` with: - `result.content.html` — full HTML string - `result.content.markdown` — markdown version - `result.content.cleaned_html` — cleaned HTML - `result.content.readability` — readability text - `result.metadata.status_code` — HTTP status (int) - `result.metadata.title` — page title - `result.links` — list of extracted links - Always release sessions when done: `client.sessions.release(sessionId)` - Do NOT use `session.websocketUrl` directly — construct the WSS URL as shown above - The Node SDK constructor param is `steelAPIKey` (not `apiKey`) - The Python SDK constructor param is `steel_api_key` (not `api_key`) - Individual doc pages are available as markdown at `/llms.mdx/` ## Pages - [Cookbook](https://docs.steel.dev/cookbook): Runnable recipes for using Steel with your favorite libraries and frameworks. - [Integrations](https://docs.steel.dev/integrations): Connect Steel with browser agents, AI frameworks, and automation tools — including Playwright, Puppeteer, Stagehand, Browser Use, OpenAI Agents, and Claude Computer Use. #### Overview - [Sessions Api Overview](https://docs.steel.dev/overview/sessions-api/overview): The Sessions API lets you create and control cloud-based browser sessions through simple API calls. Each session is like a fresh incognito window, but running in our cloud and controlled through code. - [Profiles Api Overview](https://docs.steel.dev/overview/profiles-api/overview): Reuse browser context, auth, cookies, extensions, credentials, and browser settings across sessions. - [Files Api Overview](https://docs.steel.dev/overview/files-api/overview): How to upload, download, manage and work with files within an active session - [Extensions Api Overview](https://docs.steel.dev/overview/extensions-api/overview): Add Chrome extensions to your Steel sessions. - [Credentials API](https://docs.steel.dev/overview/credentials-api/overview): Programmatic access to storing credentials for users or agents. - [Captchas Api Overview](https://docs.steel.dev/overview/captchas-api/overview): Automatically detect and solve CAPTCHAs in browser sessions using Steel's integrated captcha solvers and the CAPTCHAs API. - [Browser Tools](https://docs.steel.dev/overview/browser-tools/overview): One-shot endpoints for scraping pages, capturing screenshots, and generating PDFs without managing a session yourself. - [Agent Traces: Observability for Browser Agents](https://docs.steel.dev/overview/agent-traces/overview): Agent Traces adds observability to browser agents: review recorded Steel sessions as a readable timeline synced with video, and export for debugging and replay. - [Agent Traces API](https://docs.steel.dev/overview/agent-traces/api): Fetch the activity timeline for a Steel browser session as JSON. - [Authentication](https://docs.steel.dev/overview/authentication): Authenticate requests to the Steel API and Steel SDKs using an API key. - [Available Skills](https://docs.steel.dev/overview/skills/available-skills): Choose the Steel Skill that matches your agent workflow. - [CAPTCHA Solving API for Browser Automation](https://docs.steel.dev/overview/stealth/captcha-solving): Solve CAPTCHAs automatically in Steel browser sessions with an API built for AI agents. Handles checkbox, invisible, image, and slider challenges. - [Clustering](https://docs.steel.dev/overview/self-hosting/clustering): Self-Hosting a Steel Browser Cluster - [Dedicated IPs](https://docs.steel.dev/overview/sessions-api/dedicated-ips): Give Steel sessions a stable network identity with dedicated IPs, and combine them with profiles for durable browser identity. - [Docker](https://docs.steel.dev/overview/self-hosting/docker): Self-Hosting Steel Browser Using Docker - [Embed Sessions](https://docs.steel.dev/overview/sessions-api/embed-sessions): Learn how to view and embed your live and past sessions. - [Extensions](https://docs.steel.dev/overview/self-hosting/extensions): Self-Hosting Steel Browser Using Extensions - [Fullscreen Mode](https://docs.steel.dev/overview/sessions-api/fullscreen-mode): Launch browser sessions edge-to-edge with no Chrome UI, so pages get the full 1920×1080 and screenshots stay clean. - [Implement Human-in-the-Loop Controls](https://docs.steel.dev/overview/sessions-api/human-in-the-loop): How to let users take control of Steel browser sessions - [Intro to Steel](https://docs.steel.dev/overview/intro-to-steel): Humans use Chrome, Agents use Steel. - [Legal](https://docs.steel.dev/overview/legal): This page outlines the legal terms and conditions for using Steel. - [Live Sessions](https://docs.steel.dev/overview/sessions-api/embed-sessions/live-sessions): How to embed and share live browser sessions in your applications - [Mobile Mode](https://docs.steel.dev/overview/sessions-api/mobile-mode): Create browser sessions that appear as mobile devices with full mobile fingerprints and touch capabilities. - [Multi-region](https://docs.steel.dev/overview/sessions-api/multi-region): Control where your Steel browser sessions are hosted for optimal performance and latency. - [Need Help?](https://docs.steel.dev/overview/need-help): Need help with Steel? Check out our documentation or reach out to use on Discord. - [Overview Getting Started](https://docs.steel.dev/overview/skills): Install Steel Skills so coding agents can use Steel cloud browsers correctly. - [Past Sessions](https://docs.steel.dev/overview/sessions-api/embed-sessions/past-sessions): How to access recordings of past browser sessions and display them within your app - [Pricing/Limits](https://docs.steel.dev/overview/pricinglimits): Steel pricing for Launch, Scale, and Enterprise plans, including usage credits, metered rates, and plan limits. - [Profiles](https://docs.steel.dev/overview/self-hosting/profiles): Self-Hosting Steel Browser Using Profiles - [Proxies & Proxy Rotation for Browser Automation](https://docs.steel.dev/overview/stealth/proxies): Use residential proxies and proxy rotation in Steel browser sessions to avoid rate limits, access geo-restricted content, and stay undetected at scale. - [Railway](https://docs.steel.dev/overview/self-hosting/railway): A quick guide on deploying Steel Browser to Railway using our template - [Reusing Context & Auth](https://docs.steel.dev/overview/sessions-api/reusing-auth-context): How to Reuse Authentication Across Steel Sessions - [Session Lifecycle](https://docs.steel.dev/overview/sessions-api/session-lifecycle): Learn how to start and release browser sessions programmatically. - [Sessions Api Quickstart](https://docs.steel.dev/overview/sessions-api/quickstart): Get up a running with your first Steel Session in a few minutes. - [Steel Browser Skill](https://docs.steel.dev/overview/skills/available-skills/steel-browser): Use Steel Browser when an agent should operate a live Steel browser now. - [Steel CLI](https://docs.steel.dev/overview/steel-cli): Run browser automation from the terminal with the Steel CLI. Manage sessions, scrape, screenshot, and install Steel Skills for coding agents like Claude Code, Cursor, Codex, and OpenCode. - [Steel Developer Skill](https://docs.steel.dev/overview/skills/available-skills/steel-developer): Use Steel Developer when an agent should write reusable Steel code, scripts, examples, or app integrations. - [Steel Local vs Steel Cloud](https://docs.steel.dev/overview/self-hosting/steel-local-vs-steel-cloud): What's the difference between local Steel and Steel Cloud? - [Steel Reliability Skill](https://docs.steel.dev/overview/skills/available-skills/steel-reliability): Use Steel Reliability for bot-detection, CAPTCHA, proxy, identity, login persistence, pacing, and retry strategy. - [Steel Session Debugging Skill](https://docs.steel.dev/overview/skills/available-skills/steel-session-debugging): Use Steel Session Debugging to diagnose failed Steel browser sessions from logs, traces, replay artifacts, and metadata. - [Steel Skill Creator](https://docs.steel.dev/overview/skills/available-skills/steel-skill-creator): Use Steel Skill Creator to turn recurring browser workflows into reusable agent workflows. - [Stripe Projects](https://docs.steel.dev/overview/stripe-projects): Provision a Steel project and STEEL_API_KEY from the Stripe CLI with `stripe projects add steel/browser`. Credentials sync into .env so your agents can start cloud browser sessions. - [Timeline and exports](https://docs.steel.dev/overview/agent-traces/timeline): How to read the Agent Traces timeline in the session viewer, sync it with the video, and export the run as markdown, JSON, or a ZIP bundle. - [Troubleshooting Skills](https://docs.steel.dev/overview/skills/troubleshooting): Fix common Steel Skills install, auth, and routing issues. - [WebRTC](https://docs.steel.dev/overview/self-hosting/webrtc): Self-Hosting Steel Browser Using WebRTC #### Integrations - [AgentKit](https://docs.steel.dev/integrations/agentkit): Build multi-agent networks in TypeScript with routing, shared state, and MCP tools. - [Agno](https://docs.steel.dev/integrations/agno): Build Python agent teams with shared memory, reasoning, and RAG. - [Browser Tool for the Vercel AI SDK](https://docs.steel.dev/integrations/ai-sdk): Give the Vercel AI SDK a browser tool with Steel. Build typed TypeScript agents that stream across any model provider and browse the real web. - [Build a CrewAI Browser Agent on Steel](https://docs.steel.dev/integrations/crewai): Give CrewAI agents a real browser with Steel. Orchestrate multi-agent Python crews that navigate the web, with shared memory and human-in-the-loop reviews. - [Build a LangGraph Browser Agent on Steel](https://docs.steel.dev/integrations/langgraph): Build a LangGraph browser agent on Steel cloud browsers. Model web tasks as explicit state machines with nodes, edges, and structured output in Python. - [Build a Pydantic AI Browser Agent on Steel](https://docs.steel.dev/integrations/pydantic-ai): Build a Pydantic AI browser agent on Steel cloud browsers: provider-agnostic, typed Python agents with dependency injection and real web access. - [Claude Agent SDK + Steel Cloud Browsers](https://docs.steel.dev/integrations/claude-agent-sdk): Use the Claude Agent SDK with a Steel cloud browser: build web-browsing agents on Anthropic's first-party agent loop with in-process MCP tools. - [Claude Code](https://docs.steel.dev/integrations/claude-code): Give Claude Code a real cloud browser to drive from the terminal. - [Claude Computer Use](https://docs.steel.dev/integrations/claude-computer-use): Wire Claude's computer-use tool into a Steel browser session. - [Codex](https://docs.steel.dev/integrations/codex): Give Codex a real cloud browser to drive from the terminal. - [Connect Puppeteer to Steel Cloud Browsers (CDP)](https://docs.steel.dev/integrations/puppeteer): Run Puppeteer in the cloud with Steel. Connect over CDP in TypeScript with no local Chrome, plus built-in stealth, rotating proxies, and CAPTCHA solving. - [Gemini Computer Use](https://docs.steel.dev/integrations/gemini-computer-use): Wire Gemini 3's computer-use tool into a Steel browser session. - [Hermes Agent](https://docs.steel.dev/integrations/hermes-agent): Pending integration that adds Steel as a browser provider inside Hermes. - [Magnitude](https://docs.steel.dev/integrations/magnitude): Turn natural-language prompts into typed browser actions and structured outputs. - [Mastra](https://docs.steel.dev/integrations/mastra): Build typed TypeScript browser agents with Mastra's Model Router, registry, and Studio playground. - [Notte](https://docs.steel.dev/integrations/notte): Reliable Python web-agent framework focused on stable navigation and structured outputs. - [OpenAI Agents SDK + Steel Cloud Browsers](https://docs.steel.dev/integrations/openai-agents-sdk): Use the OpenAI Agents SDK with a Steel cloud browser. Build typed browser agents with handoffs, guardrails, and tracing on managed infrastructure. - [OpenAI Computer Use](https://docs.steel.dev/integrations/openai-computer-use): Wire OpenAI's computer-use tool into a Steel browser session. - [OpenClaw](https://docs.steel.dev/integrations/openclaw): Give OpenClaw a real cloud browser for form-heavy workflows from the terminal. - [Pi Agent](https://docs.steel.dev/integrations/pi-agent): Drop-in extension that gives Pi a real cloud browser and structured extract. - [Replit](https://docs.steel.dev/integrations/replit): Remix Steel browser-automation starter templates and run them in Replit's cloud. - [Run Browser Use in the Cloud on Steel](https://docs.steel.dev/integrations/browser-use): Run Browser Use in the cloud with Steel. Point the open-source Python agent at a Steel session and let LLMs navigate, fill forms, and extract data at scale. - [Run Playwright on Steel Cloud Browsers](https://docs.steel.dev/integrations/playwright): Run Playwright in the cloud with Steel. Connect over CDP in TypeScript or Python and get managed browsers with stealth, rotating proxies, and CAPTCHA solving. - [Run Selenium in the Cloud on Steel](https://docs.steel.dev/integrations/selenium): Run Selenium in the cloud with Steel instead of running your own grid. Stealth, rotating proxies, and CAPTCHA solving built in, over W3C WebDriver in Python. - [Run Stagehand on Steel Cloud Browsers](https://docs.steel.dev/integrations/stagehand): Run Stagehand on Steel cloud browsers. Write automations in natural language with act, extract, and observe. Stealth and proxies are managed for you. - [StackBlitz / Bolt.new](https://docs.steel.dev/integrations/stackblitz-bolt.new): Run Steel TypeScript starter templates in your browser, no install. Bolt.new ready. - [x402](https://docs.steel.dev/integrations/x402): Pay-per-use Steel API calls with USDC on Base and Solana. No API key, no account. #### Cookbook - [Agents](https://docs.steel.dev/cookbook/topics/agents): Agent frameworks that run a perception-plan-act loop against a Steel browser session. - [Authentication](https://docs.steel.dev/cookbook/topics/authentication): Patterns for persisting and replaying authenticated sessions across Steel browsers. - [Automate a cloud browser with chromedp](https://docs.steel.dev/cookbook/chromedp): Use Steel with chromedp to connect over CDP, navigate to Hacker News, extract the top stories, and capture a screenshot. - [Automate a cloud browser with chromiumoxide](https://docs.steel.dev/cookbook/chromiumoxide): Use Steel with chromiumoxide to connect over CDP, drive the handler task, extract page content, and capture a screenshot. - [Automate a cloud browser with headless_chrome](https://docs.steel.dev/cookbook/headless-chrome): Use Steel with headless_chrome, the synchronous Rust equivalent of Puppeteer, to connect over CDP and scrape quotes with element handles. - [Automate a cloud browser with Playwright](https://docs.steel.dev/cookbook/playwright): Use Steel with Playwright in TypeScript for cloud browser automation. - [Automate a cloud browser with Puppeteer](https://docs.steel.dev/cookbook/puppeteer): Use Steel with Puppeteer in TypeScript for cloud browser automation. - [Automate a cloud browser with Rod](https://docs.steel.dev/cookbook/rod): Use Steel with Rod's fluent, chainable API to connect over CDP and scrape quotes.toscrape.com from a cloud browser. - [Automate a cloud browser with Selenium](https://docs.steel.dev/cookbook/selenium): Use Steel with Selenium in Python for cloud browser automation. - [Automate browsing with natural-language instructions using Stagehand](https://docs.steel.dev/cookbook/stagehand): Use Steel with Stagehand for natural-language-driven AI browser automation. - [Automate logins with the Credentials API](https://docs.steel.dev/cookbook/credentials): Use the Steel Credentials API with Playwright to automate flows with stored credentials. - [Brian Sparker](https://docs.steel.dev/cookbook/authors/bsparker): 1 recipe contributed to the Steel Cookbook by Brian Sparker. - [Browser automation](https://docs.steel.dev/cookbook/topics/browser-automation): Drive a cloud browser with familiar automation libraries: Playwright, Puppeteer, Selenium, Stagehand. - [Browser Use](https://docs.steel.dev/cookbook/topics/browser-use): Agent recipes built on the browser-use framework. - [Build a browser agent with Agno](https://docs.steel.dev/cookbook/agno): Integrate Steel with the Agno agent framework. - [Build a browser agent with Browser Use](https://docs.steel.dev/cookbook/browser-use): Integrate Steel with the browser-use framework for AI-driven web automation. - [Build a browser agent with Eino](https://docs.steel.dev/cookbook/eino): Use Steel with the ByteDance Eino framework to build a ReAct agent that calls Steel's scrape API as a tool to research and answer a web question. - [Build a browser agent with Genkit](https://docs.steel.dev/cookbook/genkit): Use Steel with Genkit Go to build a tool-calling agent that navigates and extracts from a chromedp-backed browser and completes a web task. - [Build a browser agent with Google ADK](https://docs.steel.dev/cookbook/google-adk): Use Steel with Google's Agent Development Kit (ADK) for Go to build a tool-using browser agent that drives a chromedp session over CDP and reads Hacker News. - [Build a browser agent with Inngest AgentKit](https://docs.steel.dev/cookbook/agentkit): Integrate Steel with Inngest's AgentKit framework. - [Build a browser agent with LangChainGo](https://docs.steel.dev/cookbook/langchaingo): Use Steel with LangChainGo's zero-shot ReAct (MRKL) agent and a string-in, string-out scrape tool so Claude reads a page and answers a question. - [Build a browser agent with Microsoft Agent Framework](https://docs.steel.dev/cookbook/microsoft-agent-framework): Use Steel with Microsoft Agent Framework 1.0 (the successor to AutoGen and Semantic Kernel) to build a tool-using browser agent. - [Build a browser agent with rig](https://docs.steel.dev/cookbook/rig): Use Steel with rig to build an agent that drives a cloud browser over CDP with chromiumoxide through navigate and extract tools, then answers a multi-step web task. - [Build a browser agent with the Claude Agent SDK](https://docs.steel.dev/cookbook/claude-agent-sdk): Use Steel with the Claude Agent SDK (TypeScript) to build a tool-using browser agent on Anthropic's first-party agent loop. - [Build a multi-agent browser workflow with CrewAI](https://docs.steel.dev/cookbook/crewai): Integrate Steel with the CrewAI multi-agent framework. - [Build a research agent with Swiftide](https://docs.steel.dev/cookbook/swiftide): Use Steel with Swiftide to build an agent whose tool reads the web through Steel's scrape endpoint, so the model works from clean Markdown with no browser library. - [Build a typed browser agent with LangGraph](https://docs.steel.dev/cookbook/langgraph): Use Steel with LangGraph to build a typed browser agent with an explicit state-machine loop and a structured-output formatter node. - [Build a typed browser agent with Mastra](https://docs.steel.dev/cookbook/mastra): Use Steel with Mastra to build a typed browser agent with the Mastra Model Router and Studio playground. - [Build a typed browser agent with Pydantic AI](https://docs.steel.dev/cookbook/pydantic-ai): Use Steel with Pydantic AI to build typed, provider-agnostic browser agents with dependency injection. - [Build a typed browser agent with the OpenAI Agents SDK](https://docs.steel.dev/cookbook/openai-agents): Use Steel with the OpenAI Agents SDK for TypeScript to build typed, tool-using browser agents. - [Build a typed browser agent with the Vercel AI SDK](https://docs.steel.dev/cookbook/vercel-ai-sdk): Use Steel with the Vercel AI SDK v6 ToolLoopAgent for typed, tool-using browser agents. - [Build an AI browser agent with Magnitude](https://docs.steel.dev/cookbook/magnitude): Use Steel with Magnitude for AI-powered browser automation. - [Captchas](https://docs.steel.dev/cookbook/topics/captchas): Recipes that handle CAPTCHA challenges using Steel's CAPTCHA API. - [Chat with any webpage on Convex](https://docs.steel.dev/cookbook/convex-chat-with-page): Convex app that streams an AI agent's answer about any URL. The agent runs server-side with one Steel-backed scrape tool and pages through long articles via a chunked cache. - [Collin Pfeifer](https://docs.steel.dev/cookbook/authors/aspectrr): 3 recipes contributed to the Steel Cookbook by Collin Pfeifer. - [Combine You.com search with Steel browser actions](https://docs.steel.dev/cookbook/you-com-search): Pair the You.com Search and Contents APIs with a Steel cloud browser in a search-then-act LangChain agent that prefers the cheap path and only opens a session when interaction is required. - [Computer use](https://docs.steel.dev/cookbook/topics/computer-use): Model-native browser control where the LLM sees the screen and emits actions directly. - [Control a browser with Notte's reasoning engine](https://docs.steel.dev/cookbook/notte): Control browsers with AI using Steel's infrastructure and Notte's reasoning engine. - [Convex](https://docs.steel.dev/cookbook/topics/convex): Recipes that run Steel from a Convex backend. - [Dane Wilson](https://docs.steel.dev/cookbook/authors/danew): 1 recipe contributed to the Steel Cookbook by Dane Wilson. - [Deep research with Claude Agent SDK subagents](https://docs.steel.dev/cookbook/deep-research): Lead orchestrator dispatches parallel researcher subagents, each driving its own Steel browser, and synthesizes findings into a cited Markdown report. - [Drive a browser with Claude Computer Use](https://docs.steel.dev/cookbook/claude-computer-use): Connect Claude to a Steel browser session for autonomous web interactions. - [Drive a browser with Gemini Computer Use](https://docs.steel.dev/cookbook/gemini-computer-use): Connect Google's Gemini Computer Use to a Steel browser session for autonomous web interactions. - [Drive a browser with OpenAI Computer Use](https://docs.steel.dev/cookbook/openai-computer-use): Connect OpenAI's Computer Use Assistant to a Steel browser session for autonomous web interactions. - [Drive a mobile browser with Claude Computer Use](https://docs.steel.dev/cookbook/claude-computer-use-mobile): Claude Computer Use with Steel for autonomous task execution in mobile browser environments. - [Expose a Steel browser to any MCP client](https://docs.steel.dev/cookbook/mcp): Build a Model Context Protocol server in Go with the official SDK and chromedp that hands any MCP client a Steel cloud browser through explicit session-handle tools. - [Hussien Hussien](https://docs.steel.dev/cookbook/authors/hussufo): 4 recipes contributed to the Steel Cookbook by Hussien Hussien. - [Jagadesh Jai](https://docs.steel.dev/cookbook/authors/jagadeshjai): 4 recipes contributed to the Steel Cookbook by Jagadesh Jai. - [Jun Ryu](https://docs.steel.dev/cookbook/authors/junhsss): 43 recipes contributed to the Steel Cookbook by Jun Ryu. - [MCP](https://docs.steel.dev/cookbook/topics/mcp): 1 recipe tagged MCP. - [Mobile](https://docs.steel.dev/cookbook/topics/mobile): Recipes targeting Steel's mobile browser environment. - [Move files between your machine and a cloud browser](https://docs.steel.dev/cookbook/files): Use the Steel Files API with Playwright to automate file uploads and downloads in the cloud. - [Next.js](https://docs.steel.dev/cookbook/topics/nextjs): Recipes that integrate Steel into a Next.js application. - [Nikola Balic](https://docs.steel.dev/cookbook/authors/nibzard): 3 recipes contributed to the Steel Cookbook by Nikola Balic. - [Persist authenticated sessions with Profiles](https://docs.steel.dev/cookbook/profiles): Maintain authenticated sessions across Steel browser instances using profiles. - [Playwright](https://docs.steel.dev/cookbook/topics/playwright): Recipes that use Playwright to drive a Steel session, either as the automation library itself or as transport for Steel APIs. - [Restate](https://docs.steel.dev/cookbook/topics/restate): 1 recipe tagged Restate. - [Reuse authenticated sessions across browsers](https://docs.steel.dev/cookbook/auth-context): Maintain authenticated sessions across Steel browser instances by capturing and reusing cookies and local storage. - [Run a durable browser agent with Restate](https://docs.steel.dev/cookbook/restate-agent): Build a Restate Virtual Object in TypeScript that uses durable OpenAI planning steps and Steel scraping to answer browser research questions. - [Run a durable browser workflow with Temporal](https://docs.steel.dev/cookbook/temporal-browser-workflow): Build a Temporal TypeScript Workflow that schedules retryable Steel browser Activities to capture page summaries, screenshots, and Markdown artifacts. - [Run a Steel browser job with Trigger.dev](https://docs.steel.dev/cookbook/trigger-dev-browser-job): Queue a Trigger.dev task that creates a Steel session, drives Playwright over CDP, saves artifacts, and releases the browser in cleanup. - [Scrape JavaScript-Rendered Pages to Markdown](https://docs.steel.dev/cookbook/scrape): Scrape JavaScript-rendered pages to clean Markdown for LLM context with the Steel API, plus screenshots and PDFs. No browser library required. - [Search](https://docs.steel.dev/cookbook/topics/search): Recipes that pair a search API with a Steel browser to keep the agent on the cheap path and only open a session when interaction is required. - [Solve CAPTCHAs automatically in a Browser Use agent](https://docs.steel.dev/cookbook/browser-use-captcha-auto): Build an AI agent with browser-use and Steel that solves CAPTCHAs automatically. - [Solve reCAPTCHA v2 manually with Browser Use](https://docs.steel.dev/cookbook/browser-use-captcha-manual): Manually solve reCAPTCHA v2 using Steel's CAPTCHA API with the browser-use framework. - [Steel APIs](https://docs.steel.dev/cookbook/topics/steel-apis): Recipes for Steel's first-party APIs: credentials, auth contexts, profiles, extensions, files. - [Stream a browser agent into a Next.js chat app](https://docs.steel.dev/cookbook/vercel-ai-sdk-nextjs): A Next.js App Router chat app where a Vercel AI SDK agent drives a Steel cloud browser with embedded Live View. - [Subagents](https://docs.steel.dev/cookbook/topics/subagents): Multi-agent recipes where a lead orchestrator dispatches parallel subagents, each driving its own Steel browser. - [Temporal](https://docs.steel.dev/cookbook/topics/temporal): 1 recipe tagged Temporal. - [Trigger.dev](https://docs.steel.dev/cookbook/topics/triggerdev): 1 recipe tagged Trigger.dev. - [Typed output](https://docs.steel.dev/cookbook/topics/typed-output): Agents that return structured, schema-validated results instead of free-form text. - [Upload and run browser extensions](https://docs.steel.dev/cookbook/extensions): Use the Steel Extensions API with Playwright to upload and run browser extensions. - [Watch Claude pricing for divergent A/B variants](https://docs.steel.dev/cookbook/convex-price-watch): Convex cron plus two parallel Steel proxy probes against claude.com/pricing. Stores per-tier per-region snapshots and surfaces tiers where the probes disagree. #### Changelog - [Changelog #022](https://docs.steel.dev/changelog/changelog-022): Agent-native Steel Cloud onboarding, AI SDK and OpenAI Agents examples, the beta x402 pay-per-use endpoint, and CLI shell completion. - [Changelog #023](https://docs.steel.dev/changelog/changelog-023): New Claude Agent SDK, PydanticAI, Mastra, and LangGraph integrations, plus Steel CLI forge improvements and infra fixes. - [Changelog #024](https://docs.steel.dev/changelog/changelog-024): Deep research cookbook examples using Claude Agent SDK subagents, plus a round of Files API docs corrections. - [Changelog #025](https://docs.steel.dev/changelog/changelog-025): Fullscreen browser sessions, new cookbook examples, a click-to-interact viewer overlay, and live viewer and WebRTC playback fixes. - [Changelog #026](https://docs.steel.dev/changelog/changelog-026): Agent traces in the session viewer, bearer-token API auth, broader stealth coverage, and a new browser tools docs section. - [Changelog #027](https://docs.steel.dev/changelog/changelog-027): A dedicated Agent Traces docs section, a rebuilt leaderboard results browser, and llms.txt delivery fixes. - [Changelog #028](https://docs.steel.dev/changelog/changelog-028) - [Changelog #029](https://docs.steel.dev/changelog/changelog-029) - [Changelog #030](https://docs.steel.dev/changelog/changelog-030) - [Changelog #031](https://docs.steel.dev/changelog/changelog-031) - [Changelog #032](https://docs.steel.dev/changelog/changelog-032) - [Changelog #033](https://docs.steel.dev/changelog/changelog-033) - [Changelog #034](https://docs.steel.dev/changelog/changelog-034) - [Changelog #035](https://docs.steel.dev/changelog/changelog-035)