Skip to main content
voiceai is a single static binary that wraps the SLNG API. Use it to generate audio, transcribe files, stream over WebSocket, pipe audio between tools, and manage agents and other resources from the command line.

​
Install

Verify the install:
On macOS, the curl-installed binary is unsigned. If Gatekeeper blocks the first run, run xattr -d com.apple.quarantine $(which voiceai), or right-click the binary in Finder and choose Open once.

​
Authenticate

The quickest way to get started is voiceai login. It walks you through the setup step by step and supports multiple profiles, so you can keep separate keys for different accounts or environments.
Run it again to add another profile. Create a key at app.slng.ai, or see Create your API key. For non-interactive setup, such as CI, export the key in your shell instead:
The CLI reads its config from ~/.config/voiceai/config.json. Environment variables take precedence over a saved profile.

​
Text to speech

Synthesize audio from text. By default, the CLI plays the audio.
Pipe raw audio when stdout is not a terminal:
Stream chunks over WebSocket for long text or low-latency playback:

​
Speech to text

Transcribe a file:
Stream from the microphone:
Pipe raw PCM from arecord or any source:

​
Discover models and voices

​
Manage agents and resources

Beyond speech, the CLI manages the resources in your account: voice agents, projects, and the rest. Run voiceai --help for the full command list, and each subcommand’s --help for its options. To learn what the agent commands do, see Create an agent and Outbound calls.

​
Configuration

Defaults the CLI reads from ~/.config/voiceai/config.json: Set any of them with voiceai config set <key> <value>. Override per invocation with the matching flag.

​
Source and releases