---
title: FAQ
description: Get answers to the most frequently asked questions about AgentGPT
icon: "question"
---

Discover answers to the most frequently asked questions about AgentGPT. If your question isn't covered here, feel free to connect with our team on [Discord](https://discord.gg/jdSBAnmdnY) or [GitHub](https://github.com/reworkd/AgentGPT).

## Agent issues

<AccordionGroup>

<Accordion title="AgentGPT said it made a file / database / script, where can I find it?">
  At present, AgentGPT does not have the capability to generate outputs in this
  way. However, this is a feature we're actively developing. To keep updated on
  this, check out our [roadmap](/essentials/roadmap).
</Accordion>

<Accordion title="Did AgentGPT really email 100 people?">
  Not yet! We're looking into implementing this functionality. To see when this
  might be available, take a look at our [roadmap](/essentials/roadmap). We guarantee that
  actions such as "sending an email" will only occur with your explicit
  permission.
</Accordion>

<Accordion title="My agent keeps running out of loops!">
  Due to limitations on API usage and infrastructure costs, there's a cap on how
  much the Agent can run. To circumvent this, consider hosting AgentGPT locally
  using your own API key. Alternatively, subscribing to our pro plan will allow
  for increased limits. More details can be found in our [setup
  documentation](/essentials/roadmap).
</Accordion>

<Accordion title="My Agent was running smoothly, but then it stopped! Can I resume from where it left off?">
  Currently, each Agent run is independent, so resuming isn't possible. However,
  you can initiate another AgentGPT run using information from the previous run.
  We plan on adding this feature in the future. Stay updated via our
  [roadmap](/essentials/roadmap).
</Accordion>

<Accordion title="Can AgentGPT use GPT-4?">
  Currently, free-tier users of AgentGPT utilize GPT-3.5, while PRO users have
  access to GPT-4. If you have API access to GPT-4, you can run AgentGPT locally
  using your API key.
</Accordion>

<Accordion title="Why does the output keep getting cut off?">
  The output length is limited to manage generation costs on our end. If you're
  using your own API key, you can adjust the output length within the advanced
  settings by increasing the token count. If you're a Pro user or are a running
  AgentGPT locally, you can adjust the output length winthin the advanced
  settings
</Accordion>

</AccordionGroup>

## Misc

<AccordionGroup>

<Accordion title="How is AgentGPT different from ChatGPT?">
  While ChatGPT is designed to answer specific questions and maintain a
  conversation, AgentGPT is a platform where you can configure an AI agent to
  achieve a broad goal.
</Accordion>

<Accordion title="How is AgentGPT different from AutoGPT?">
  Both AgentGPT and AutoGPT are autonomous AI agent projects. However, while
  AutoGPT runs locally, AgentGPT is a web-based platform.
</Accordion>

<Accordion title="What can AgentGPT do?">
  AgentGPT boasts a wide array of capabilities and we're always working to
  expand them. To see some of its applications, visit our [usecases](/essentials/usecases)
  page.
</Accordion>

<Accordion title="Can AgentGPT access the internet?">
  Yes, AgentGPT does have internet access.
</Accordion>

</AccordionGroup>

## Local contribution issues

<AccordionGroup>

<Accordion title="I'm experiencing issues setting up AgentGPT locally!">
  For troubleshooting, refer to our [setup guide](/development/setup). If your
  issue persists, you're welcome to submit an [issue on
  GitHub](https://github.com/reworkd/AgentGPT/issues).
</Accordion>

</AccordionGroup>