This is a CLI for Hume AI's Octave TTS API.
Unlike conventional TTS that merely "reads" words, Octave is a speech-language model that understands what words mean in context, unlocking a new level of expressiveness. It acts out characters, generates voices from prompts, and takes instructions to modify the emotion and style of a given utterance.
This CLI uses Hume's Typescript SDK behind the scenes.
npm install -g @humeai-cli
hume login
# Use the browser to login to platform.hume.ai to retrieve your
# API keys
hume tts "Are you serious?" --description "whispered, hushed"
hume voices create --name whisperer --last
hume tts "I said, are you serious?" --voice-name whisperer
hume voices list # View your saved voices
hume voices list --provider HUME_AI # View Hume's voice library
hume voices delete --name whisperer # Delete a voice when no longer neededThe Hume CLI is distributed via NPM. You can install it globally via:
npm install -g @humeai/cliText to speech
━━━ Usage ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
$ hume tts <text>
━━━ Options ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
-d,--description #0 Description of the desired voice
-c,--continue,--context-generation-id #0 Previous generation ID for continuation
-l,--last,--continue-from-last Use a generation from a previous synthesis as context. If the last synthesis was created with --num-generations > 1, you must also provide --last-index
--last-index #0 Index of the generation to use from the previous synthesis.
-o,--output-dir #0 Output directory for generated audio files
-n,--num-generations #0 Number of variations to generate
-p,--prefix #0 Filename prefix for generated audio
--play #0 Play audio after generation: all variations, just the first, or none
--play-command #0 Command to play audio files (uses $AUDIO_FILE as placeholder for file path)
--api-key #0 Override the default API key
--format #0 Output audio format
-v,--voice-name #0 Name of a previously saved voice
--voice-id #0 Direct voice ID to use
--json Output in JSON format
--pretty Output in human-readable format
--base-url #0 Override the default API base URL (for testing purposes)
--provider #0 Voice provider type (CUSTOM_VOICE or HUME_AI)
--speed #0 Speaking speed multiplier (0.25-3.0, default is 1.0)
--trailing-silence #0 Seconds of silence to add at the end (0.0-5.0, default is 0.35)
--streaming Use streaming mode for TTS generation (default: true)
--instant-mode Enable ultra-low latency mode for significantly faster generation (requires streaming=true, a voice, and incurs 10% higher cost)
━━━ Details ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
This command converts text to speech using Hume AI's advanced AI voice
synthesis. You can specify voice characteristics through descriptions or use
saved voices.
━━━ Examples ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Basic usage
$ hume tts "Make sure to like and subscribe!" --description "The speaker is a charismatic, enthusiastic, male YouTuber in his 20s with a American accent, a slightly breathy voice, and a fast speaking rate."
Saving a voice you like (see `hume voices create --help`)
$ hume voices create --name influencer_1 --last
Using a previously-saved voice
$ hume tts "Thanks for the 100,000,000,000 likes guys!" -v influencer_1
Using a voice from the Hume Voice Library
$ hume tts "Hello there" -v narrator --provider HUME_AI
Reading from stdin
$ echo "I wouldn't be here without you" | hume tts - -v influencer_1
Continuing previous text
$ hume tts "Take some arrows from the quiver" -v influencer_1
$ hume tts "Take a bow, too" -v influencer_1 --last # should rhyme with 'toe' not 'cow'
Using custom audio player (macOS/Linux)
$ hume tts "Hello world" -v narrator --play-command "mpv $AUDIO_FILE --no-video"
Using custom audio player (Windows)
$ hume tts "Hello world" -v narrator --play-command "powershell -c \"[System.Media.SoundPlayer]::new('$AUDIO_FILE').PlaySync()\""
Setting a custom audio player for the session
$ hume session set tts.playCommand "vlc $AUDIO_FILE --play-and-exit"
Adjusting speech speed
$ hume tts "I am speaking very slowly" -v narrator --speed 0.75
Adding trailing silence
$ hume tts "Wait for it..." -v narrator --trailing-silence 3.5
Using instant mode for ultra-low latency
$ hume tts "Hello world" -v narrator --instant-mode
Setting instant mode in your config (always enable)
$ hume config set tts.instantMode true
## Voice Management
Save a voice from a previous generation
$ hume voices create --name my-narrator --last
Create a voice from a specific generation ID
$ hume voices create --name my-narrator --generation-id abc123
### Listing Voices
List your custom voices
$ hume voices list
List voices from the Hume Voice Library
$ hume voices list --provider HUME_AI
### Deleting Voices
Delete a voice by name
$ hume voices delete --name my-narrator
━━━ See also ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
- `hume voices create --help` - Save a voice for later use
- `hume voices list --help` - List available voices
- `hume voices delete --help` - Delete a saved voice
- `hume session --help` - Save settings temporarily so you don't have to repeat yourself
- `hume config --help` - Save settings more permanently