Back to vendors
I

Intuned

Also known as: Intuned Agent, intunedhq.com, intuned.io

Visit site
Entry price$25 a month; Startup $120; Free plan; Enterprise customFull pricing detail

Intuned is a code first browser automation platform where an AI agent writes, tests, deploys and repairs Playwright scrapers, crawlers and RPA workflows. The agent explores the target site in a real browser, proposes a plan for approval, builds on its own branch and validates the result on production infrastructure. The hosted runtime runs the automations on schedules or by API with saved logins, stealth, proxies and captcha solving, and can fix them when sites change.

Intuned, a Y Combinator company, sells a platform for building and running browser automations as code: scrapers, crawlers and RPA workflows written in Playwright with TypeScript or Python. Developers can write the code themselves or describe what they need to Intuned Agent, and the hosted runtime runs the result in production with concurrency limits, retries and no browser infrastructure to manage. Fiber AI runs thousands of automations a day on it, and GovSignals and a GovTech company's chief technology officer describe building and maintaining hundreds of scrapers there.

Intuned Agent works inside the platform with a real browser and the Intuned CLI. It explores the target site, proposes a plan that a developer can change or approve, writes the code on its own agent branch, tests it on the same infrastructure used in production and runs a full validation before anyone merges. Developers inspect the code and test artifacts, then merge and deploy or merge only.

The agent also repairs projects from failed runs, fixing broken selectors, changed page structure, timeouts and wrong output, and reads Playwright traces to find the cause. With self healing switched on, it watches a project's runs, writes fixes and redeploys without waiting for a person. Code is deterministic by default, with optional AI driven steps through Stagehand or Browser-Use, and a private preview plugin runs the agent inside a developer's own Claude Code or Codex.

The runtime handles the hard parts of production browsing. Auth sessions created from a developer's own scripts are validated, reused and recreated, stealth mode, proxies and captcha solving help on protected sites, and jobs run scheduled or batched executions while the run API starts single executions on demand. Results flow to webhooks, Amazon S3, Cloudflare R2 or Supabase, and every run keeps logs, traces, session recordings and saved records. Enterprise adds SAML or OpenID Connect sign in, storage in the customer's own cloud, OpenSearch monitoring with alerts to Slack, email or a webhook, custom regions and desktop automation for Windows and Linux applications.

The Free plan includes 3 compute hours and 10 job runs, Developer costs $25 a month with 100 compute hours, Startup costs $120 a month with 500 compute hours, and Enterprise is custom. AI credits are included on each plan, or teams can use their own Anthropic or OpenAI key. Intuned shows SOC 2 and HIPAA compliance through its trust center. It suits engineering teams that need dependable scrapers and web automations against sites without APIs, while teams wanting a no code tool for business users will find a developer platform.

Vendor details

Canonical URL

https://intunedhq.com/

Category

Browser / computer-use agent

Subcategory

AI agent that builds and runs Playwright browser automations

Company status

independent

Use cases & customers

Primary use cases

Build and maintain web scrapers and crawlers with an AI agentRun RPA workflows on sites without APIsRepair automations automatically when websites change

Target customers

Engineering and data teamsStartups building on web dataGovernment technology and compliance teams automating portals

Deployment options

Intuned hosted browser runtimeBring your own storage in the customer's cloud on EnterpriseLocal development through the CLI

Integrations

Playwright projects in TypeScript or Python, run and job APIs, sinks to webhooks, Amazon S3, Cloudflare R2 and Supabase, Stagehand and Browser-Use for AI driven steps, a CLI for CI/CD, and Anthropic or OpenAI keys for AI usage.

In practice

A data team asks the agent for a scraper of a state procurement portal, approves its plan and merges the tested project.

A nightly job logs into a supplier portal with a saved auth session, downloads invoices and drops them into S3.

When a retailer changes its product page layout, self healing rewrites the selectors and redeploys before the next run.

Agentic Index coverage score

11.5 / 14 capabilities · 82%

Integrations & Tool Calling Full

Automations act on websites with saved, validated logins, filling forms, submitting and downloading as RPA workflows require, and their results flow out through sinks to webhooks, Amazon S3, Cloudflare R2 and Supabase with documented retry behavior. Developers can add their own integrations in code, since every project is a full Playwright project in TypeScript or Python running on Intuned.

Sourceintunedhq.com/docs/llms.txtread 2026-11-20

Workflow Orchestration Full

Jobs run batched executions across many inputs, with payloads that can expand during a run to queue more work, concurrency limits and retries, while runs handle single executions. The agent works in stages of its own: it explores the site, plans, builds, tests on production infrastructure and validates before handing back a project, and it can repair failed runs in the same loop.

Sourceintunedhq.com/docs/main/02-intuned-agent/overview.mdread 2026-11-20

Knowledge Grounding & RAG Partial

The agent grounds each task in the live target site, which it explores in a real browser, and in the project's own results, traces and logs when it repairs a failed run. It matches the language and structure of existing code when editing. Context is gathered for each session, and loading project instructions and custom skills is still marked as coming soon.

Sourceintunedhq.com/docs/main/02-intuned-agent/overview.mdread 2026-11-20

Human Oversight & Guardrails Full

The agent proposes a plan that a developer can send back with suggested changes or approve, then works on its own agent branch. After the build the developer reviews the code and test artifacts and chooses to merge and deploy, merge only or keep iterating. Self healing, which redeploys fixes without review, is a setting each project turns on deliberately.

Sourceintunedhq.com/docs/main/02-intuned-agent/overview.mdread 2026-11-20

Security, Identity & Governance Full

Intuned shows SOC 2 security controls and HIPAA compliance through a trust center. Workspaces manage users and API keys, and Enterprise adds single sign on through the customer's identity provider over SAML 2.0 or OpenID Connect along with advanced access controls. Teams can use their own Anthropic or OpenAI key so agent prompts run under their own provider agreement.

Sourceintunedhq.com/docs/main/02-features/enterprise/overview.mdread 2026-11-20

Observability & Auditability Full

Every run keeps logs, Playwright traces, session recordings and saved records, so a developer can replay what an automation did on a page and see why it failed. The agent reads the same traces when it repairs a project, so its fixes start from the evidence. Enterprise adds OpenSearch dashboards, reports and alerts across runs and job runs, sent to Slack, email or a webhook.

Sourceintunedhq.com/docs/main/02-features/enterprise/overview.mdread 2026-11-20

Memory & State Persistence Not documented

Auth sessions keep validated logins that the runtime reuses and recreates, which is browser state the platform applies rather than context the agent reads. The agent's code, branches and run history persist in the project. Agent memory through project instructions and custom skills is announced as coming soon, so no memory layer the agent consults across sessions is available yet.

Sourceintunedhq.com/docs/main/02-intuned-agent/overview.mdread 2026-11-20

Deployment & Data Residency Full

Automations run on Intuned's hosted browser runtime, and Enterprise customers can keep run data in their own cloud account, with an S3 compatible bucket and a DynamoDB table for the persistent store. Enterprise plans also offer custom regions, retention and compliance setup. Developers build and test locally through the CLI before deploying to the hosted runtime.

Sourceintunedhq.com/docs/main/02-features/enterprise/overview.mdread 2026-11-20

Prebuilt Agents / Templates / Packs Partial

A public cookbook on GitHub and quickstarts for scrapers and RPA give developers working examples to start from, and the agent produces a complete project from a plain description. Those examples become the team's own code once adopted. Intuned keeps no catalog of ready made automations for common sites that teams can subscribe to and run unchanged.

Sourceintunedhq.com/docs/llms.txtread 2026-11-20

Triggers & Channel Coverage Full

Jobs run on schedules or when triggered through the API, and the run API starts single executions on demand from an application or pipeline, so data refreshes arrive without anyone pressing a button. The CLI fits deployments into CI/CD. With self healing switched on, failed runs themselves start the agent, which detects the issue and writes a fix without anyone opening the dashboard.

Sourceintunedhq.com/docs/main/02-intuned-agent/overview.mdread 2026-11-20

Model Flexibility & Routing Partial

Intuned chooses the models behind its agent and bills them as token based AI credits. Teams can configure their own Anthropic or OpenAI key so AI requests run on their account instead, though the agent itself needs an Anthropic key. AI driven steps inside an automation can run through Stagehand or Browser-Use, which the developer configures in code.

Sourceintunedhq.com/pricingread 2026-11-20

APIs / SDKs / MCP Extensibility Full

Client APIs start, read and cancel runs, create, pause and trigger jobs, list job runs, manage auth sessions and configure sinks, with a full API reference for each call. Runtime SDKs for TypeScript and Python expose helpers for auth sessions, captchas, tracing and structured data extraction, and a CLI handles local development, configuration and deployment.

Sourceintunedhq.com/docs/llms.txtread 2026-11-20

Testing, Debugging & Optimization Full

Before handing back a project, the agent tests its code on the same infrastructure used in production and runs a full validation, and the developer reviews the results in an artifacts panel before merging. After deployment, self healing watches runs, detects broken selectors, timeouts or wrong output and puts a fix through the same build and test loop.

Sourceintunedhq.com/docs/main/02-intuned-agent/overview.mdread 2026-11-20

Browser / Computer-use Full

Intuned runs real browsers that it hosts, with stealth mode, proxies and captcha solving, and the agent explores target sites in a real browser before writing Playwright code that clicks, types and navigates on its own. Optional AI driven steps use Stagehand or Browser-Use, and Enterprise adds unattended desktop automation for Windows and Linux applications.

Sourceintunedhq.com/docs/main/02-intuned-agent/overview.mdread 2026-11-20

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

Pricing

$25 a month; Startup $120; Free plan; Enterprise custom

Monthly plan with included compute hours, job runs and AI credit, plus overages

Free tierTrial available

Included quota

Free includes 3 compute hours, 5 API machines, 10 job runs and a $5 trial AI credit with 7 day retention. Developer includes 100 compute hours, 25 API machines and 100 job runs with 30 day retention. Startup includes 500 compute hours, 100 API machines, 400 job runs and $40 a month of AI credit.

What is public

Intuned's Free plan costs $0, Developer costs $25 a month and Startup costs $120 a month, each with included compute hours, API machines, job runs and AI credit. Standalone runs are unlimited on every plan.

Billing mechanics

Compute hours, job runs and larger machines are metered beyond each plan's allowance. AI agent usage draws on token based AI credits, and a team that configures its own Anthropic or OpenAI key pays that provider instead, while other metering still follows the plan.

Overage / add-ons

Developer overages cost $0.12 per compute hour and $0.12 per job run, and Startup overages $0.10 each. Large machines add $0.18 or $0.15 an hour and extra large machines $0.48 or $0.40 an hour.

Sales call required

Mixed (some tiers require a call)

Free / trial

Free plan; trials of paid plans

Lowest paid plan

Developer

Commercial notes

Enterprise adds custom compute and machine limits, volume pricing for runs and AI, single sign on, advanced access controls, custom retention, regions and compliance setup, dedicated onboarding and a dedicated support team.

Missing data

Intuned gives two figures for the Developer plan's monthly AI credit, $10 and $5. Enterprise prices come from sales.

Agentic Index verified 2026-11-20

Contact us

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