PrankGPT
prankgpt.com
2
Leaving SiteNav
External Link Disclaimer
You are about to visit prankgpt.com. This website is not operated by us. We are not responsible for its content or privacy practices.
About this website
PrankGPT is a web-based application that allows users to generate automated prank phone calls using AI-powered voice agents. The service is built on top of the Vocode open-source library, which specializes in creating voice-based interactions using large language models (LLMs). Users initiate a prank by entering the phone number of their target recipient, selecting one of two distinct voice personas—Marv, described as the original evil prankbot, or Zephyr, a Gen Z queen voice provided by rime.ai—and then typing a prompt or instruction that guides the AI on what topics to discuss during the call. Once the call is placed, the AI engages in a real-time conversation with the person on the other end, following the user’s predefined script and reacting naturally to responses. The platform leverages Google Cloud’s text-to-speech and speech-to-text capabilities to handle voice synthesis and recognition, while Rime Labs supplies additional voice assets for enhanced variety and expressiveness. The entire call flow is orchestrated by Vocode, which manages the state of the conversation, turns speech into text, processes it through an LLM, converts the generated text back into speech, and delivers it over a phone network. This architecture enables calls to feel spontaneous and adaptive, even though they are fully automated. Common use cases for PrankGPT include lighthearted practical jokes among friends, testing someone’s reaction to absurd scenarios, or simply entertaining a group by setting up unexpected but harmless conversations. For instance, a user might instruct the AI to pretend to be a confused delivery driver looking for an address, or to ask ridiculous survey questions about fictional products. The two voice options offer contrasting tones—Marv delivers a darker, more mis
Statistics
2
Views
0
Clicks
0
Like
0
Dislike