Skip to content

Meet

Stop building workflows. Start hiring agents.

Press Enter to send, Shift+Enter for new line

As featured in

    Cited & referenced by

      Logos are property of their respective publishers.

      Core features

      Built for the way you actually work. Chat, build, oversee.

      Three surfaces, one platform. Describe what you want, fine-tune the details, and keep an eye on everything that's running — all without switching tools.

      Automate by chatting. Skip the setup.

      Tell AutoPilot what you want done, the way you'd brief a teammate. It learns your process, builds the agent, and puts the work on autopilot — all without leaving the conversation.

      • Go from idea to working agent in one chat.
      • No flowcharts, prompt-engineering, or API setup.
      • Built-in expertise picks the right tools for you.
      • Refine in plain English whenever something changes.
      • Your process, captured the moment you describe it.
      Try AutoPilot
      AutoPilot — the chat-driven automation interface in AutoGPT

      Integrations

      Connect what's yours. We've wired up the rest.

      Every model, built in. No API keys, no setup — just connect the tools that are already yours and your agents get to work.

      • Chat & reasoning
      • Image generation
      • Video generation
      • Text-to-speech
      • Transcription
      • Music
      • Image editing
      • Vision
      • OCR
      • 3D
      • Embeddings

      Already connected — no API key needed

        Bring the accounts that are yours

          45+ platforms · hundreds of AI models · zero API keys

          In their words

          What people are saying about AutoGPT

          Founders, researchers, and builders on what AutoGPT unlocked.

          Built for every role

          One platform. Every team's busywork.

          Executive ops

          Your morning brief, already written.

          Earnings, headlines, internal pulse — pulled, ranked, and on your desk before standup. No dashboards to babysit.

          Build my morning brief
          AutoGPT generating an executive daily briefing

          Sales

          Walk in knowing exactly who you're meeting.

          Account history, recent news, last-touch context, and the three questions worth asking — assembled for every meeting on tomorrow's calendar.

          Prep my next call
          AutoGPT preparing a sales call briefing

          Marketing

          From brief to full campaign draft overnight.

          Positioning, channel cuts, subject lines, and a launch checklist — drafted while you sleep, ready for your edits in the morning.

          Draft my campaign
          AutoGPT drafting a multi-channel marketing campaign

          Finance

          Markets move. You hear about it first.

          Tickers, filings, and sector signals watched continuously. You only get pinged when something actually matters to your book.

          Watch my markets
          AutoGPT delivering a market watch update

          Engineering

          Incidents triaged before you open Slack.

          Logs cross-referenced, recent deploys checked, likely culprit named. Your on-call starts with a hypothesis, not a blank terminal.

          Triage my incidents
          AutoGPT analyzing an engineering incident

          Customer support

          Every ticket, already half-answered.

          Drafted replies, pulled order history, flagged escalations — your team reviews and sends instead of typing from scratch.

          Clear my queue
          AutoGPT triaging a customer support queue

          Pricing

          Simple, transparent pricing

          Upgrade, downgrade, or change plans anytime. All plans include core features to get you started.

          Pro

          1x usage
          $42.50 / month
          Charged today: $510.00

          For individuals getting started with dependable everyday automation.

          Access to virtually any leading AI model
          Agents work non-stop in the background
          Visual agent builder and chat-based agent creation
          File-aware agents that can work with your documents
          End-to-end agent management and run visibility
          Scheduled and event-based triggers
          Best value

          Max

          8.5x usage
          $272.00 / month
          Charged today: $3,264.00

          For users who are serious about getting more work done.

          Includes everything in Pro
          8.5x Pro usage
          5x file storage
          Early access to features before anyone else
          Expand our extensive integration library
          Priority support
          Help drive the roadmap for new features

          Team

          Coming soon
          Contact us

          For teams that need collaboration, controls, and custom usage.

          Multi-user workspaces
          Admin controls & roles
          Centralized billing & seat management
          Collaboration & approvals
          Security & compliance options
          Secure access for shared integrations
          See full pricing & compare every feature

          FAQ

          Questions, answered.

          What is AutoGPT?

          An AI agent is software you can hand a goal to — it makes its own plan, then carries out the steps to reach it, using AI models and your apps along the way. AutoGPT is the platform for building and running those agents. Describe what you want in plain language or assemble it visually, and your agents get to work — no code required.

          How does AutoGPT work?

          You create an agent by describing the outcome you want in plain English, or by wiring up the steps in the visual builder. AutoGPT connects to the AI models and apps the job needs, runs the work in the background, and shows you every step of every run. Set it to run on a schedule or fire on a trigger, and it keeps working without you watching.

          What can I build with it?

          Anything that's repeatable work rather than a one-off: monitoring the web and alerting you when something changes, drafting and sending outreach, researching prospects before a meeting, turning long documents into summaries, keeping spreadsheets and CRMs up to date. If you can describe the job, you can usually build an agent for it.

          How does AutoGPT compare to hiring a contractor?

          A contractor is the right call for judgement-heavy, one-off work. AutoGPT is built for the repeatable kind: an agent runs around the clock, handles many tasks at once, responds in seconds, and costs a fraction of an hourly rate. Most people use both — a person for the bespoke work, agents for everything that happens again and again.

          How is AutoGPT different from Claude Code, Codex, or Perplexity?

          Tools like Claude Code and Codex help you write code, and Perplexity answers questions with sources — they assist you, in the moment, while you're prompting them. AutoGPT is for building agents that run on their own: across your apps, on a schedule or trigger, doing the whole job end to end rather than the part you're actively working on.

          Can I use AutoGPT for free?

          Yes, if you're technical: AutoGPT is open source, so you can self-host the whole platform for free on your own infrastructure (you bring your own model API keys). The hosted cloud doesn't have a free tier — every agent run uses real compute and paid AI-model usage that we cover — so instead each plan includes enough to put agents to work from day one.

          What are credits, and what counts as usage?

          Running agents is pay-as-you-go: each automation draws from a pre-paid credit wallet you can top up any time, and the rate is the same on every plan. Your subscription (Pro or Max) sets how much AutoPilot Chat usage is included — that's the main difference between plans. In short: the subscription covers chat, and the credit wallet covers the agents you run. See the pricing page for the full breakdown.

          Can I self-host AutoGPT?

          Yes. AutoGPT is open source, and you can run the entire platform yourself with a single Docker Compose command — on Linux, WSL, Podman, even a Raspberry Pi. Self-hosting is free; you bring your own model API keys and infrastructure. The hosted cloud is the managed, zero-setup option if you'd rather not run servers. Start from the GitHub repo or the docs.

          Still have a question? Read the docs or ask in Discord.