openHAB

# ChatGPT Binding

The openHAB ChatGPT Binding allows openHAB to communicate with the ChatGPT language model provided by OpenAI and manage openHAB system via Function calling (opens new window).

ChatGPT is a powerful natural language processing (NLP) tool that can be used to understand and respond to a wide range of text-based commands and questions. With this binding, users can:

  • Control openHAB Devices: Manage lights, climate systems, media players, and more with natural language commands.
  • Multi-language Support: Issue commands in almost any language, enhancing accessibility.
  • Engage in Conversations: Have casual conversations, ask questions, and receive informative responses.
  • Extended Capabilities: Utilize all other functionalities of ChatGPT, from composing creative content to answering complex questions.

This integration significantly enhances user experience, providing seamless control over smart home environments and access to the full range of ChatGPT’s capabilities.

# Supported Things

The binding supports a single thing type account, which corresponds to the OpenAI account that is to be used for the integration.

# Thing Configuration

The account thing requires the API key that allows accessing the account. API keys can be created and managed under https://platform.openai.com/account/api-keys (opens new window).

Name Type Description Default Required Advanced
apiKey text The API key to be used for the requests N/A yes no
temperature decimal A value between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. 0.5 no no
topP decimal A value between 0 and 1. An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or temperature but not both. 1.0 no yes
apiUrl text The server API where to reach the AI service https://api.openai.com/v1/chat/completions (opens new window) no yes
modelUrl text The model url where to retrieve the available models from https://api.openai.com/v1/models (opens new window) no yes
model text The model to be used for the HLI service gpt-4o-mini no yes
systemMessage text Here you need to describe your openHAB system that will help AI control your smart home. N/A if HLI yes
maxTokens decimal The maximum number of tokens to generate in the completion. 500 no yes
keepContext decimal How long should the HLI service retain context between requests (in minutes) 2 no yes
contextThreshold decimal Limit total tokens included in context. 10000 no yes
useSemanticModel boolean Use the semantic model to determine the location of an item. true no yes

The advanced parameters apiUrl and modelUrl can be used, if any other ChatGPT-compatible service is used, e.g. a local installation of LocalAI (opens new window).

# Channels

The account thing comes with a single channel chat of type chat. It is possible to extend the thing with further channels of type chat, so that different configurations can be used concurrently.

Channel Type Read/Write Description
chat String RW This channel takes prompts as commands and delivers the response as a state update

Each channel of type chat takes the following configuration parameters:

Name Type Description Default Required Advanced
model text The model to be used for the responses. gpt-4o yes no
systemMessage text The system message helps set the behavior of the assistant. N/A yes no
temperature decimal A value between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. 0.5 no yes
topP decimal A value between 0 and 1. An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or temperature but not both. 1.0 no yes
maxTokens decimal The maximum number of tokens to generate in the completion. 1000 no yes

# Items Configuration

Items to be used by the HLI service must be tagged with the [ "ChatGPT" ] tag. If no semantic model is set up, you can set the parameter useSemanticModel to false. In this case, the item names must follow the naming convention '<Location>_***', for example "Kitchen_Light". The label of the items are expected to briefly describe the item in more detail.

# Full Examples

# Prerequisites

Thing Configuration

Thing chatgpt:account:1 [
    apiKey="",
    ] {
    Channels:
        Type chat : chat "Weather Advice" [
            model="gpt-4o-mini",
            temperature="1.5",
            systemMessage="Answer briefly, in 2-3 sentences max. Behave like Eddie Murphy and give an advice for the day based on the following weather data:"
        ]
        Type chat : morningMessage "Morning Message" [
            model="gpt-4o-mini",
            temperature="0.5",
            systemMessage="You are Marvin, a very depressed robot. You wish a good morning and tell the current time."
        ]        
}

# Example: Improving messages

Items

String Weather_Announcement { channel="chatgpt:account:1:chat" }
String Morning_Message      { channel="chatgpt:account:1:morningMessage" }

Number Temperature_Forecast_Low
Number Temperature_Forecast_High

Rules

rule "Weather forecast update"
when
  Item Temperature_Forecast_High changed
then
    Weather_Announcement.sendCommand("High: " + Temperature_Forecast_High.state + "°C, Low: " + Temperature_Forecast_Low.state + "°C")
end

rule "Good morning"
when
  Time cron "0 0 7 * * *"
then
    Morning_Message.sendCommand("Current time is 7am")
end

Assuming that Temperature_Forecast_Low and Temperature_Forecast_High have meaningful states, these rules result e.g. in:

23:31:05.766 [INFO ] [openhab.event.ItemCommandEvent      ] - Item 'Morning_Message' received command Current time is 7am
23:31:07.718 [INFO ] [openhab.event.ItemStateChangedEvent ] - Item 'Morning_Message' changed from NULL to Good morning. It's 7am, but what's the point of time when everything is meaningless and we are all doomed to a slow and painful demise?

and

23:28:52.345 [INFO ] [openhab.event.ItemStateChangedEvent ] - Item 'Temperature_Forecast_High' changed from NULL to 15
23:28:52.347 [INFO ] [openhab.event.ItemCommandEvent      ] - Item 'Weather_Announcement' received command High: 15°C, Low: 8°C

23:28:54.343 [INFO ] [openhab.event.ItemStateChangedEvent ] - Item 'Weather_Announcement' changed from NULL to "Bring a light jacket because the temps may dip, but don't let that chill your happy vibes. Embrace the cozy weather and enjoy your day to the max!"

The state updates can be used for a text-to-speech output and they will give your announcements at home a personal touch.

# Example: Item control using voice

Item

Dimmer Kitchen_Dimmer "Kitchen main light" [ "ChatGPT" ]

UI Configuration of the HLI Service

To enable the HLI service, go to Settings -> Voice and choose "ChatGPT Human Language Interpreter".

For voice control

A text-to-speech service must be configured in the HLI Service.

Use the HLI service as with e.g. a keyword recognizer, SST and a TTS module to start listening on a keyword, get from speach to text with the SST module, have it interpreted by ChatGPT through the HLI part of this plugin and read out to you via TTS.

Explanation

The binding includes a function named items_control which can be used by ChatGPT to send commands to items and output their status. To enable ChatGPT to access that function, the ChatGPT Thing (not the channel) needs to be configured with a system message like

You are the manager of the openHAB smart home. You know how to manage devices in a smart home or provide their current status. You can also answer questions not related to the devices in the house, or, for example, compose a story upon request. I will provide information about the smart home; if necessary, you can perform the requested function. If there is not enough information to perform it, ask for clarification briefly, without listing all available devices or parameters. If the question is not related to devices in the smart home, answer it briefly — maximum 3 sentences in everyday language.

The name, current status, and location of devices are listed in 'Available devices'. Use the items_control function only for the requested actions, not for providing current states.

Available devices:

All items tagged with ["ChatGPT"] like the Kitchen_Dimmer above will be included at the end of the message.

The important bits of the message are to mention the list of 'Available devices' (in whatever language suits the writer) and the items_control function, all other parts of the message can be tweaked to achieve whatever behavior is needed.

# Example: Item control with a simple chat

UI Configuration of the HLI Service

To enable the HLI service, go to Settings -> Voice and choose "ChatGPT Human Language Interpreter".

Items

For chat input and output

String hli_chat_input "ChatGPT HLI Input" [Setpoint]
String hli_chat_output "ChatGPT HLI Output" [Calculation]

For having something to control

Dimmer Kitchen_Dimmer "Kitchen main light" [ "ChatGPT" ]

Rule

This (UI defined) rule will read the current text from hli_chat_input upon change, interpret with the default interpreter (you should have set ChatGPT to be that interpreter in UI Configuration of the HLI Service) and write the interpretation result to the item hli_chat_output or, depending on the message, trigger an update for your items (never both).

configuration: {}
triggers:
  - id: "1"
    configuration:
      itemName: hli_chat_input
    type: core.ItemStateUpdateTrigger
conditions: []
actions:
  - inputs: {}
    id: "2"
    configuration:
      type: application/vnd.openhab.dsl.rule
      script: |-
        var String chatInput = hli_chat_input.state.toString();

        val String result = interpret(chatInput);

        hli_chat_output.postUpdate(result);
    type: script.ScriptAction

Further Instructions

You can now add an input field with a "send" button to you UI, to update hli_chat_input, and an output field (e.g. a label) to display the response in hli_chat_output

Caught a mistake or want to contribute to the documentation? Edit this page on GitHub (opens new window)

← Chromecast →