Merge branch 'features/Realtime-WebRTC'

Author: ngelx

README.md

@@ -67,6 +67,7 @@ Stream chats with the Responses API, transcribe and translate audio with Whisper
       - [Translate](#translate)
       - [Transcribe](#transcribe)
       - [Speech](#speech)
+    - [Real-Time](#real-time)
     - [Usage](#usage)
     - [Errors](#errors-1)
   - [Development](#development)
@@ -1587,6 +1588,33 @@ File.binwrite('demo.mp3', response)
 # => mp3 file that plays: "This is a speech test!"
 ```
 
+### Real-Time
+
+The Real-Time API allows you to create a real-time session with an OpenAI model. It responds with a session object, plus a client_secret key which contains a usable ephemeral API token that can be used to authenticate browser clients for the Realtime API.
+
+```ruby
+response = client.real_time.create(parameters: { model: "gpt-4o-realtime-preview-2024-12-17" })
+puts "ephemeral key: #{response.dig('client_secret', 'value')}"
+# => "ephemeral key: ek_abc123"
+```
+
+Then in the client-side application, make a POST request to the Real-Time API with the ephemeral key and the SDP offer.
+
+```js
+const OPENAI_REALTIME_URL = 'https://api.openai.com/v1/realtime/sessions'
+const MODEL = 'gpt-4o-realtime-preview-2024-12-17'
+
+const response = await fetch(`${OPENAI_REALTIME_URL}?model=${MODEL}`, {
+  method: 'POST',
+  headers: {
+    'Content-Type': 'application/sdp',
+    'Authorization': `Bearer ${ephemeralKey}`,
+    'OpenAI-Beta': 'realtime=v1'
+  },
+  body: offer.sdp
+})
+```
+
 ### Usage
 The Usage API provides information about the cost of various OpenAI services within your organization.
 To use Admin APIs like Usage, you need to set an OPENAI_ADMIN_TOKEN, which can be generated [here](https://platform.openai.com/settings/organization/admin-keys).

lib/openai.rb

@@ -10,6 +10,7 @@
 require_relative "openai/assistants"
 require_relative "openai/threads"
 require_relative "openai/messages"
+require_relative "openai/real_time"
 require_relative "openai/runs"
 require_relative "openai/run_steps"
 require_relative "openai/vector_stores"

lib/openai/client.rb

@@ -92,6 +92,10 @@ def batches
       @batches ||= OpenAI::Batches.new(client: self)
     end
 
+    def real_time
+      @real_time ||= OpenAI::RealTime.new(client: self)
+    end
+
     def moderations(parameters: {})
       json_post(path: "/moderations", parameters: parameters)
     end

lib/openai/real_time.rb

@@ -0,0 +1,25 @@
+module OpenAI
+  class RealTime
+
+    DEFAULT_REALTIME_MODEL = 'gpt-4o-realtime-preview-2024-12-17'
+
+    def initialize(client:)
+      @client = client.beta(realtime: 'v1')
+    end
+
+    # Create a new real-time session with OpenAI.
+    #
+    # This method sets up a new session for real-time voice interaction with an OpenAI model.
+    # It returns session details that can be used to establish a WebRTC connection.
+    #
+    # By default, this method uses the 'gpt-4o-realtime-preview-2024-12-17' model unless specified otherwise.
+    #
+    # @param parameters [Hash] parameters for the session (see: https://platform.openai.com/docs/api-reference/realtime-sessions/create)
+    # @return [Hash] Session details including session ID, ICE servers, and other connection information
+    def create(parameters: {})
+      parameters = parameters.merge(model: DEFAULT_REALTIME_MODEL) unless parameters[:model]
+
+      @client.json_post(path: '/realtime/sessions', parameters: parameters)
+    end
+  end
+end