Back to vendors
S

Steel

Also known as: Steel Browser, Steel.dev

Visit site
Entry priceLaunch $0 plus usage · Scale $250/mo plus usage · browser hours from $0.08Full pricing detail

Open source browser infrastructure for AI agents, with cloud or self hosted Chromium sessions, stealth, CAPTCHA solving, persistent profiles and step by step agent traces.

Steel is open source browser infrastructure for AI agents. It runs isolated cloud Chromium sessions an agent drives to navigate, click, fill forms and read pages, through a REST API, Node and Python SDKs, a CLI, or CDP connections for Playwright, Puppeteer and Selenium. The core runtime, Steel Browser, is open source and self-hosts with Docker, a one-click Railway deploy or a clustered setup. The managed cloud runs browsers in a single region, us-east.

Steel is built for production web work. Profiles persist cookies, local storage, extensions and auth context across sessions, and a Credentials API stores logins with envelope encryption and injects them into sessions, with TOTP support, without letting the agent read the secrets back. Proxies, stealth settings, dedicated IPs, mobile mode and CAPTCHA solving keep automations from being blocked.

For debugging and oversight, Agent Traces turns each recorded session into a timeline of the agent's actions synced to video, exportable as markdown or JSON and available through the API, alongside console and network logs. An embeddable live view lets a person take control of a running session. Installable skills teach coding agents to use Steel, debug failed sessions and harden reliability, and a cookbook covers dozens of frameworks and languages.

Pricing is usage-based on top of three plans. Launch is $0 plus usage with $30 of one-time credits. Scale is $250 a month plus usage with $100 of monthly credits, Enterprise SSO and a HIPAA-ready BAA. Custom Enterprise plans come with reserved browser pools and sessions up to 24 hours. Browser hours run $0.10 on Launch and $0.08 on Scale, billed by the minute, and an inactivity timeout releases idle sessions.

Vendor details

Canonical URL

https://steel.dev

Category

Agent infrastructure

Subcategory

Browser infrastructure

Funding status

Independent. The core runtime, Steel Browser, is open source (github.com/steel-dev/steel-browser) and self-hostable, alongside a managed cloud.

Company status

independent

Use cases & customers

Primary use cases

cloud browsers for AI agentsauthenticated web access with stored credentials and profilesscraping and extraction on protected sitesdebugging and auditing browser agent runs

Target customers

developersAI agent buildersweb automation teams

Deployment options

SaaSself-hosted

Integrations

REST API with official Node and Python SDKs and a CLI, plus CDP connections for Playwright, Puppeteer and Selenium. Integrations give agents a Steel browser inside Vercel AI SDK, LangGraph, CrewAI, Pydantic AI, Mastra, OpenAI Agents SDK, Claude Agent SDK, Browser Use, Stagehand, Hermes Agent and Claude, OpenAI and Gemini computer use. The Credentials API injects stored logins, installable agent skills serve coding agents, and scrape, screenshot and PDF tools turn pages into HTML, markdown and images.

In practice

Your agent needs to browse real websites, not brittle scrapes. You create a Steel session over the API, connect Playwright via Chrome DevTools Protocol, and the agent navigates, clicks, and extracts with anti detection handled.

An agent keeps getting lost mid task and you cannot tell why. Steel's live Session Viewer with replays lets you watch the exact page state it saw, turning debugging into inspection instead of guessing.

Compliance requires keeping browsing data private. You self host the open source Steel container in your own environment, so sessions and cookies never leave your infrastructure.

Agentic Index coverage score

7.0 / 14 capabilities · 50%

Integrations & Tool Calling Full

Agents reach outside tools and systems through Steel in two ways. A named integration catalog gives the agent a Steel browser as a tool inside more than twenty frameworks and agents (Vercel AI SDK, LangGraph, CrewAI, Pydantic AI, Mastra, OpenAI Agents SDK, Claude Agent SDK, Browser Use, Stagehand, Hermes Agent and others). The Credentials API lets the agent act in outside systems under stored logins. Credentials are stored with envelope encryption, injected into a session's logins with TOTP support, and isolated so the agent cannot read them back.

Sourcedocs.steel.dev/integrationsread 2026-09-23

Workflow Orchestration Not documented

Each browser session has a lifecycle Steel manages. Sessions are created, kept alive, and released on an inactivity timeout or an explicit call. Steel does not run multi step or multi agent workflows itself. Multi step recipes in the cookbook run in the customer's chosen engine (CrewAI, LangGraph, Temporal, Restate, Convex), and Steel has no workflow runtime or step chaining of its own.

Sourcedocs.steel.dev/overview/sessions-api/session-lifecycleread 2026-09-23

Knowledge Grounding & RAG Not documented

Retrieval over the customer's own content is not part of the product. Steel names no knowledge base, index, vector store or grounding layer of its own. Its scrape, screenshot and PDF tools turn live web pages into HTML, markdown and images for the customer's own pipeline. They extract from the public web and do not search the customer's documents.

Sourcedocs.steel.dev/overview/browser-tools/overviewread 2026-09-23

Human Oversight & Guardrails Partial

A person can step into a running session. Embedding a session's live view with interactive=true and showControls=true lets a person take control of the browser to enter credentials, solve a CAPTCHA, or verify and correct what the agent did. Steel names no approval gate, pause primitive, action allow list or policy guardrail. The customer's own code decides when to hand control to a person.

Sourcedocs.steel.dev/overview/sessions-api/human-in-the-loopread 2026-09-23

Security, Identity & Governance Partial

Enterprise SSO through the customer's identity provider comes on the Scale and Enterprise plans. The Credentials API protects each credential with its own AES-256-GCM key wrapped by an organization specific KMS key, and keeps secrets out of the agent's reach after injection. Scale and Enterprise also offer a HIPAA ready BAA. Steel publishes no trust center, security page, certification badge or attestation such as a SOC 2 report. SSO comes as a plan feature, and Steel names no roles or permissions.

Sourcedocs.steel.dev/overview/pricinglimitsread 2026-09-23

Observability & Auditability Full

Agent Traces turns each recorded session into a timeline of agent activity, one row per action with verb, target, page URL and timestamp, expandable to page, element, pointer, keyboard or error detail. The timeline is synced to the session recording and built from what happened in the browser, not from what the agent claimed.

Traces export as markdown, JSON or a ZIP with screenshots, and a REST endpoint serves them. Sessions carry console and network logs, and data retention runs by plan (7 days on Launch, 14 on Scale, custom on Enterprise). Steel does not say whether it captures the agent's prompts, and names no audit log separate from the traces.

Sourcedocs.steel.dev/overview/agent-traces/overviewread 2026-09-23

Memory & State Persistence Not documented

Between runs, Steel keeps the browser's state. Profiles persist cookies, local storage, extensions and auth context across sessions. Session context can be captured from one run and replayed into the next, and the Credentials API stores logins for injection. The browser loads each of these at session start, and the agent does not read them as context for its decisions. Steel names no memory the agent reads, and no scope or lifetime for one.

Sourcedocs.steel.dev/overview/profiles-api/overviewread 2026-09-23

Deployment & Data Residency Full

Customers can run Steel in their own environment. The open source Steel Browser runtime and API self host in the customer's infrastructure with Docker, a one click Railway deploy, or a clustered setup, with guides for each. The managed cloud runs browsers in one fixed region, us-east, and rejects any other region value. A customer who needs a different hosting location gets it only by self hosting. Proxy geolocation changes the IP a site sees, not where the browser runs.

Sourcedocs.steel.dev/overview/self-hosting/dockerread 2026-09-23

Prebuilt Agents, Templates & Packs Partial

Five installable agent skills for coding agents (steel-browser, steel-developer, steel-reliability, steel-session-debugging and a skill creator) come with Steel, along with a cookbook of starter projects across frameworks and languages. The skills teach the customer's own agent to use Steel. The starters are examples a developer builds from. Steel offers no catalog of agents a customer adopts to do work.

Sourcedocs.steel.dev/overview/skills/available-skillsread 2026-09-23

Triggers & Channel Coverage Not documented

A browser starts only when a caller creates a session through the API, SDK or CLI. Steel has no event, schedule, webhook or inbound queue of its own that starts agent work. Scheduled recipes in the cookbook run on the customer's scheduler (Convex crons, Temporal workflows), not on Steel.

Sourcedocs.steel.dev/overview/sessions-api/overviewread 2026-09-23

Model Flexibility & Routing Not documented

The model lives in the customer's agent, whether Claude, OpenAI or Gemini computer use or a framework the customer runs. In each integration, that agent calls a Steel browser. Model choice and routing sit with the customer's agent, and Steel runs no model of its own for the customer to choose.

Sourcedocs.steel.dev/integrationsread 2026-09-23

APIs, SDKs & MCP Extensibility Full

Outside systems call Steel through a REST API at api.steel.dev, authenticated by API key. Its reference covers sessions, profiles, files, extensions, credentials, CAPTCHAs, browser tools and agent traces. Steel also offers official Node and Python SDKs, a CLI, and CDP connections for Playwright, Puppeteer and Selenium. The open source Steel Browser exposes the same API when self hosted.

Sourcedocs.steel.dev/api-referenceread 2026-09-23

Testing, Debugging & Optimization Partial

The Session Debugging skill has a coding agent collect evidence on a failed session and diagnose it. It hands off to the Reliability skill when the fix is bot detection, CAPTCHA, proxy, profile or retry strategy. Steel names no evaluation harness, scored test cases or quality gate.

Sourcedocs.steel.dev/overview/skills/available-skills/steel-session-debuggingread 2026-09-23

Browser & Computer Use Full

Agents drive isolated cloud browser sessions from Steel to navigate, click, fill forms and read pages. They work through CDP with Playwright, Puppeteer or Selenium, or as Claude, OpenAI and Gemini computer use agents. Operating a real browser for agents is Steel's core product, with mobile mode, stealth, proxies, CAPTCHA solving and persistent profiles for production use.

Sourcedocs.steel.dev/overview/intro-to-steelread 2026-09-23

The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Product updates

·IntegrationsVerified

Steel released a native browser plugin for Hermes Agent. The integration provides Hermes with cloud browser sessions, live session viewing, page extraction capabilities, and built-in controls for per-site proxies and CAPTCHA solving.

Bears on: Integrations

View source
·IntegrationsVerified

Steel Browser is now available within Stripe Projects. The integration provides agents with on-demand cloud browser sessions for executing live web tasks directly within the Stripe ecosystem.

Bears on: Integrations

View source
View all 2 Steel updates →Tracked since Aug 2026 · Verified from public vendor sources

Pricing

Launch $0 plus usage · Scale $250/mo plus usage · browser hours from $0.08

Metered usage: browser hours billed by the minute, proxy bandwidth per GB, CAPTCHA solves and browser tools calls per thousand, drawn first from plan credits

Free tier

Included quota

Open source self host is free (you pay only your own infrastructure). The managed cloud gives free credits to start, then Launch, Scale, and Enterprise plans meter browser session usage against plan credits and rates, with an inactivity timeout that releases idle sessions.

What is public

Plan prices, metered rates, the free cloud credits and the open source self host path are public; Enterprise rates are custom.

Billing mechanics

Free open source self host, or managed cloud plans (Launch, Scale, Enterprise) that meter browser session time and usage credits, with an inactivity timeout to stop billing idle or crashed sessions.

Cost watchouts

Proxies and CAPTCHA solving on Launch need a $10 paid deposit before use; session length is capped by plan (15 minutes on Launch, 1 hour on Scale), and data retention is 7 or 14 days unless Enterprise. Self-hosting moves the cost to your own infrastructure.

Variable cost rationale

Cost is metered usage across browser hours, proxy bandwidth, CAPTCHA solves and browser tools, so concurrency, session length and anti-bot needs drive nearly all of it; self-hosting the open source build removes the vendor bill.

Overage / add-ons

Cloud usage meters browser session time and credits against the plan; self host has no vendor metering beyond your own infrastructure.

Sales call required

Mixed (some tiers require a call)

Free / trial

Launch at $0 a month with $30 in one-time usage credits valid 90 days; the open source Steel Browser is free to self host

Lowest paid plan

Scale $250 a month plus usage

Commercial notes

Independent open source browser API for AI agents, with a self host Docker path and a managed cloud.

Key ambiguities

Rates differ by plan and by meter (browser hours, proxy bandwidth, CAPTCHA solves, browser tools), so total cost depends on the mix a workload uses; Enterprise rates are custom.

Agentic Index verified 2026-09-23

Alternatives to Steel

The closest documented capability profiles to Steel among agent infrastructure platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.

  • Kernel8.5 / 14Adds documented Workflow Orchestration
  • Browserless8.5 / 14Adds documented Workflow Orchestration
  • Prefactor4.5 / 14A lighter documented profile than Steel
  • AgentOps5.0 / 14Adds documented Model Flexibility & Routing
  • Hyperbrowser8.0 / 14Adds documented Workflow Orchestration and Memory & State Persistence, among othersSteel vs Hyperbrowser →
  • Lakera7.0 / 14Fuller documented coverage on Human Oversight & Guardrails and Security, Identity & Governance

Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded

Head to head

Contact us

Found a vendor we missed? Have feedback on the index? We'd love to hear from you.