elevenlabs 0.0.6 → 0.0.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 22770e41ca0d3c88d2dc5f83e3e4d9de510610bf1a0adaf9bf675951a647ab30
-  data.tar.gz: 8f6ffc3ef844da02c3f45385a730ebbddfe3c711d11e1b983837153c8dbd859a
+  metadata.gz: ea563591f2116a24c1911bd964ed30bfee9dd468e9890db98a7f38ad84138bd4
+  data.tar.gz: e1c155f3dc9f5daaff7ceb283ccd7493335f639a1beb66df93adaf14ac8fac96
 SHA512:
-  metadata.gz: 1b094e808358b342f7fe8cb08cf993dbafe2bac989bcb1e4655d5de5dd2884ff83626cb098da9b7c06d60a697302d33e848419f80a26fb53f34198f7894390b7
-  data.tar.gz: 44ad5334ed45f2628a22be91a0ed307b74f7e8611f6ba71cca0b4ee79f7e510f1f655e45321e47060b413e476ed9ea9913e304a44a789ec388bb7411235d42b4
+  metadata.gz: 8559ecdf9dc45be3f7018afb389cb784603ba6c255db024f7276df60490504da366f462cbf2fb3886eeaf39f305ff530ef5ee4f9dcee577df883e62e48684af7
+  data.tar.gz: 62c761c794c8641940d5b297a0252b37dc9f41e2afcaaad777701acf7c94ee8ec3089f76559eaffd26bdf4292930cf2a9f5827f5d7ed5c523445356fa3e085b4

data/README.md

@@ -14,6 +14,8 @@ This gem provides an easy-to-use interface for:
 - **Converting text to speech** and retrieving the generated audio
 - **Designing a voice** based on a text description
 - **Streaming text-to-speech audio**
+- **Music Generation**
+- **Sound Effect Generation**
 
 All requests are handled via [Faraday](https://github.com/lostisland/faraday).
 
@@ -304,7 +306,7 @@ Designed voices cannot be used for TTS until they are created in your account.
 
 If the voice is not immediately available for TTS, wait a few seconds or check its status via client.get_voice(voice_id) until it’s "active".
 
-10. Create a multi-speaker dialogue
+11. Create a multi-speaker dialogue
 ```ruby
 inputs = [{text: "It smells like updog in here", voice_id: "TX3LPaxmHKxFdv7VOQHJ"}, {text: "What's updog?", voice_id: "RILOU7YmBhvwJGDGjNmP"}, {text: "Not much, you?", voice_id: "TX3LPaxmHKxFdv7VOQHJ"}]
 
@@ -312,6 +314,78 @@ audio_data = client.text_to_dialogue(inputs)
 File.open("what's updog.mp3", "wb") { |f| f.write(audio_data) }
 ```
 
+12. **Generate Music from prompt**
+```ruby
+audio = client.compose_music(prompt: "Lo-fi hip hop beat", music_length_ms: 30000)
+File.binwrite("lofi.mp3", audio)
+```
+
+12. **Stream Music Generated from prompt**
+```ruby
+File.open("epic_stream.mp3", "wb") do |f|
+  client.compose_music_stream(prompt: "Epic orchestral build", music_length_ms: 60000) do |chunk|
+    f.write(chunk)
+  end
+end
+```
+
+13. **Generate Music with Detailed Metadata (metadata + audio) from prompt**
+```ruby
+result = client.compose_music_detailed(prompt: "Jazz piano trio", music_length_ms: 20000)
+puts result # raw multipart data (needs parsing)
+```
+
+14. **Create a music composition plan from prompt**
+```ruby
+plan = client.create_music_plan(prompt: "Upbeat pop song with verse and chorus", music_length_ms: 60000)
+puts plan[:sections]
+```
+
+15. **Create sound effects from a prompt**
+
+Basic Usage: Simple Prompt
+  Generate a sound effect with only a text prompt, using default settings (output_format: "mp3_44100_128", duration_seconds: nil (auto-detected), prompt_influence: 0.3).
+
+```ruby
+audio_data = client.sound_generation("Futuristic laser blast in a space battle")
+
+# Save the audio to a file
+File.open("laser_blast.mp3", "wb") { |f| f.write(audio_data) }
+```
+
+Advanced Usage: Custom Duration, Influence, and Format
+Specify duration_seconds, prompt_influence, and output_format for precise control over the sound effect.
+# Generate a roaring dragon sound with specific settings
+```ruby
+audio_data = client.sound_generation(
+  "Roaring dragon in a fantasy cave",
+  duration_seconds: 3.0,
+  prompt_influence: 0.7, # Higher influence for closer adherence to the prompt
+  output_format: "mp3_22050_32"
+)
+
+# Save the audio to a file
+File.open("dragon_roar.mp3", "wb") { |f| f.write(audio_data) }
+```
+
+
+Looping Sound Effect
+Create a looping sound effect for continuous playback, such as background ambiance in a video game.
+# Generate a looping ambient sound for a haunted forest
+```ruby
+audio_data = client.sound_generation(
+  "Eerie wind and distant owl hoots in a haunted forest",
+  loop: true,
+  duration_seconds: 10.0,
+  prompt_influence: 0.5,
+  output_format: "mp3_22050_32"
+)
+# Save the audio to a file
+File.open("haunted_forest_loop.mp3", "wb") { |f| f.write(audio_data) }
+```
+
+For more details, see the ElevenLabs Sound Generation API documentation.
+
 ---
 
 ## Error Handling
@@ -368,7 +442,7 @@ gem build elevenlabs.gemspec
 Install the gem locally:
 
 ```bash
-gem install ./elevenlabs-0.0.5.gem
+gem install ./elevenlabs-0.0.8.gem
 ```
 
 ---

data/lib/elevenlabs/client.rb

@@ -9,13 +9,15 @@ module Elevenlabs
     BASE_URL = "https://api.elevenlabs.io"
 
     # Note the default param: `api_key: nil`
-    def initialize(api_key: nil)
+    def initialize(api_key: nil, open_timeout: 5, read_timeout: 120)
       # If the caller doesn’t provide an api_key, use the gem-wide config
       @api_key = api_key || Elevenlabs.configuration&.api_key
 
       @connection = Faraday.new(url: BASE_URL) do |conn|
         conn.request :url_encoded
         conn.response :raise_error
+        conn.options.open_timeout = open_timeout   # time to open connection
+        conn.options.timeout      = read_timeout   # time to wait for response
         conn.adapter Faraday.default_adapter
       end
     end
@@ -129,6 +131,48 @@ module Elevenlabs
       handle_error(e)
     end
 
+    #####################################################
+    #                 Sound Generation                  #
+    #      (POST /v1/sound-generation)                  #
+    #####################################################
+
+    # Convert text to sound effects and retrieve audio (binary data)
+    # Documentation: https://elevenlabs.io/docs/api-reference/sound-generation
+    #
+    # @param [String] text - text prompt describing the sound effect
+    # @param [Hash] options - optional parameters
+    #   :loop              => Boolean  (whether to create a looping sound effect, default: false)
+    #   :duration_seconds  => Float    (0.5 to 30 seconds, default: nil for auto-detection)
+    #   :prompt_influence  => Float    (0.0 to 1.0, default: 0.3)
+    #   :output_format     => String   (e.g., "mp3_22050_32", default: "mp3_44100_128")
+    #
+    # @return [String] The binary audio data (usually an MP3).
+    def sound_generation(text, options = {})
+      endpoint = "/v1/sound-generation"
+      request_body = { text: text }
+
+      # Add optional parameters if provided
+      request_body[:loop] = options[:loop] unless options[:loop].nil?
+      request_body[:duration_seconds] = options[:duration_seconds] if options[:duration_seconds]
+      request_body[:prompt_influence] = options[:prompt_influence] if options[:prompt_influence]
+
+      headers = default_headers
+      headers["Accept"] = "audio/mpeg"
+
+      query = {}
+      query[:output_format] = options[:output_format] if options[:output_format]
+
+      response = @connection.post("#{endpoint}?#{URI.encode_www_form(query)}") do |req|
+        req.headers = headers
+        req.body = request_body.to_json
+      end
+
+      # Returns raw binary data (often MP3)
+      response.body
+    rescue Faraday::ClientError => e
+      handle_error(e)
+    end
+
     #####################################################
     #                  Design a Voice                   #
     #      (POST /v1/text-to-voice/design)              #
@@ -405,6 +449,110 @@ module Elevenlabs
       voice_id.in?(active_voices)
     end
 
+    #####################################################
+    #                     Music API                     #
+    #####################################################
+
+    # 1. Compose music (basic)
+    # POST /v1/music
+    def compose_music(options = {})
+      endpoint = "/v1/music"
+      request_body = {
+        prompt: options[:prompt],
+        composition_plan: options[:composition_plan],
+        music_length_ms: options[:music_length_ms],
+        model_id: options[:model_id] || "music_v1"
+      }.compact
+
+      headers = default_headers.merge("Accept" => "audio/mpeg")
+      query = {}
+      query[:output_format] = options[:output_format] if options[:output_format]
+
+      response = @connection.post("#{endpoint}?#{URI.encode_www_form(query)}") do |req|
+        req.headers = headers
+        req.body = request_body.to_json
+      end
+
+      response.body # raw binary audio
+    rescue Faraday::ClientError => e
+      handle_error(e)
+    end
+
+    # 2. Stream music
+    # POST /v1/music/stream
+    def compose_music_stream(options = {}, &block)
+      endpoint = "/v1/music/stream"
+      request_body = {
+        prompt: options[:prompt],
+        composition_plan: options[:composition_plan],
+        music_length_ms: options[:music_length_ms],
+        model_id: options[:model_id] || "music_v1"
+      }.compact
+
+      headers = default_headers.merge("Accept" => "audio/mpeg")
+      query = {}
+      query[:output_format] = options[:output_format] if options[:output_format]
+
+      @connection.post("#{endpoint}?#{URI.encode_www_form(query)}") do |req|
+        req.options.on_data = Proc.new do |chunk, _|
+          block.call(chunk) if block
+        end
+        req.headers = headers
+        req.body = request_body.to_json
+      end
+
+      nil # audio streamed via block
+    rescue Faraday::ClientError => e
+      handle_error(e)
+    end
+
+    # 3. Compose detailed music (metadata + audio)
+    # POST /v1/music/detailed
+    def compose_music_detailed(options = {})
+      endpoint = "/v1/music/detailed"
+      request_body = {
+        prompt: options[:prompt],
+        composition_plan: options[:composition_plan],
+        music_length_ms: options[:music_length_ms],
+        model_id: options[:model_id] || "music_v1"
+      }.compact
+
+      headers = default_headers
+      query = {}
+      query[:output_format] = options[:output_format] if options[:output_format]
+
+      response = @connection.post("#{endpoint}?#{URI.encode_www_form(query)}") do |req|
+        req.headers = headers
+        req.body = request_body.to_json
+      end
+
+      response.body # multipart/mixed with JSON + binary audio
+    rescue Faraday::ClientError => e
+      handle_error(e)
+    end
+
+
+    # 4. Create a composition plan
+    # POST /v1/music/plan
+    def create_music_plan(options = {})
+      endpoint = "/v1/music/plan"
+      request_body = {
+        prompt: options[:prompt],
+        music_length_ms: options[:music_length_ms],
+        source_composition_plan: options[:source_composition_plan],
+        model_id: options[:model_id] || "music_v1"
+      }.compact
+
+      response = @connection.post(endpoint) do |req|
+        req.headers = default_headers
+        req.body = request_body.to_json
+      end
+
+      JSON.parse(response.body, symbolize_names: true)
+    rescue Faraday::ClientError => e
+      handle_error(e)
+    end
+
     private
 
     # Common headers needed by Elevenlabs

data/lib/elevenlabs.rb

@@ -5,7 +5,7 @@ require_relative "elevenlabs/client"
 require_relative "elevenlabs/errors"
 
 module Elevenlabs
-  VERSION = "0.0.6"
+  VERSION = "0.0.8"
 
   # Optional global configuration
   class << self

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: elevenlabs
 version: !ruby/object:Gem::Version
-  version: 0.0.6
+  version: 0.0.8
 platform: ruby
 authors:
 - hackliteracy
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-08-23 00:00:00.000000000 Z
+date: 2025-09-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: faraday
@@ -39,7 +39,8 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '1.1'
 description: This gem provides a convenient Ruby interface to the ElevenLabs TTS,
-  Voice Cloning, Voice Design, Voice dialogues and Streaming endpoints.
+  Voice Cloning, Voice Design, Voice dialogues, TTS Streaming, Music Generation and
+  Streaming endpoints.
 email:
 - hackliteracy@gmail.com
 executables: []