cd /news/large-language-models/talk-to-your-plant · home topics large-language-models article
[ARTICLE · art-36448] src=github.com ↗ pub= topic=large-language-models verified=true sentiment=↑ positive

Talk to Your Plant

OpenAI released Plant Talk, an open-source project that lets users converse with a houseplant via ChatGPT using a webcam, microphone, and the OpenAI API. The system can optionally integrate Arduino sensors for more accurate soil moisture and light readings. The project is designed for advanced builders and is licensed under Apache 2.0.

read3 min views10 publishedJun 23, 2026
Talk to Your Plant
Image: source

Plant Talk is a houseplant you can have a conversation with using ChatGPT.

Using just a computer (equipped with a webcam and a microphone) and the OpenAI API, Plant Talk gives a plant a voice. Level up the project with an Arduino or similar microcontroller and you'll be able to add more interesting data like moisture and light too. Ask how it is doing and it can check its recent observations, look at its current environment, and more.

You can get started with just a laptop, but the full version is designed for advanced builders - this is how we approached it, but this is a flexible build we encourage you to adjust based on your hardware and help from Codex.

You'll need one of each part below. You'll also need a computer with the Codex Desktop app installed, a reliable Wi-Fi connection, and an OpenAI account. If you use ChatGPT, you already have one.

  • A laptop or desktop computer
  • Google Chrome or Edge
  • Webcam
  • Microphone and speakers (built-in ones are OK!)
  • Houseplant of your choosing

For the sensing part of the system, you'll also need:

  • Arduino-compatible microcontroller with USB cable
  • Capacitive soil moisture sensor
  • LM393 light sensor module
  • Jumper wires
  • Small prototyping breadboard

Note:The system works without the Arduino. Without Arduino, Plant Talk uses just the camera to understand plant health and is much less accurate.

Feel free to read through this document to get familiar with the project. When you're ready to build, before you start wiring anything, open Codex Desktop and type:

Help me make Plant Talk https://github.com/openai/planttalk

Then complete the rest of this build directly in Codex. Codex will read this repo, take you to the first setup step, and walk you through the build from there.

Once everything is running, Plant Talk works like a small control room for your plant. Open the dashboard, take an observation, and begin a voice conversation. The system can take live sensor readings, summarize recent observations, and talk about what the plant needs.

For most users, you'll want to leave the system in Ambient Mode when in use.

Start camera: selects the webcam Plant Talk uses for observations.Observe now: sends the current camera frame for a structured plant check.Connect Arduino: reads live soil moisture and light over USB serial.Talk to plant: starts the Realtime voice conversation.Open ambient mode: switches to a full-screen conversation view.

Once the basics are working, you've got a small programmable plant interface. Make it useful, make it strange, make it fit your home, classroom, lab, or art project.

Try asking Codex to add a watering reminder, support a different sensor, change the ambient mode visuals, create a new plant persona, log long-term plant health, or adapt the project for a classroom demo. The possibilities are endless, and some starter built-in customization points include:

  • Name - change the name the voice assistant uses.
  • Personality - tune how the plant speaks and what it cares about.
  • Voice - swap the spoken voice used in conversation.
  • Observation - change what the camera analysis pays attention to.

No. You can run Plant Talk with only your computer, webcam, microphone, and OpenAI API access. But, the Arduino sensors make the readings far more useful (otherwise they're based only on what the webcam sees).

Plant Talk uses some technology not supported by all internet browsers. Chrome and Edge support it most reliably.

This project is licensed under the Apache License, Version 2.0. See LICENSE and NOTICE.

── more in #large-language-models 4 stories · sorted by recency
── more on @openai 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/talk-to-your-plant] indexed:0 read:3min 2026-06-23 ·