@byteplus/speech-cli

BytePlus Speech CLI provides command-line access to Dola Speech services, including text-to-speech, transcription, and SeedAudio generation. Use it from a terminal or an AI agent.

Requirements

Node.js 16 or later. Supports x64 and arm64 on macOS, Linux, and Windows.

Quick start

speech-cli auth login
speech-cli tts --text 'Hello from Dola Speech.' \
  --speaker '<authorized-voice-id>' --output hello.mp3

Use speech-cli auth status to check your sign-in status and speech-cli --help to see all commands.

Common commands

| Task | Command | | --- | --- | | Text-to-speech | speech-cli tts | | Audio transcription | speech-cli asr --audio-file meeting.wav | | SeedAudio 1.0 generation | speech-cli audio create --model seed-audio-1.0 | | SeedAudio 1.5 generation | speech-cli audio create --model seed-audio-1.5 | | Query a SeedAudio 1.0 task | speech-cli audio get | | Show the version | speech-cli --version |

Run a command with --help for its options, for example speech-cli audio create --help.

Install and update Skills

Install or update a specific Skill:

speech-cli skills update seedaudio15-pe

You can pass multiple names. To install or update all Skills for this product, run speech-cli skills update.

Use --skills-dir to select your agent's Skills directory. The default is ~/.agents/skills.

License

Apache-2.0. See Third-party software notice for open-source component information, or run speech-cli licenses.