CLI

Run EdgeSpeak workflows from your terminal.

Install

The CLI includes its own local speech runtime, so the desktop app does not need to be installed or running. Choose the command for your operating system.

macOS / Linux

Run this in Terminal. macOS supports Apple Silicon, and Linux supports x86_64. On Linux, the installer selects a CUDA runtime automatically for a supported NVIDIA GPU and otherwise uses CPU.

curl -fsSL https://edgespeak.com/install.sh | sh

sh ~/Downloads/install.sh

Windows

Run this in PowerShell on Windows x64. The installer uses the CPU build by default and adds edgespeak-cli to your user PATH without requiring administrator access.

irm https://edgespeak.com/install.ps1 | iex

To enable CUDA on a supported NVIDIA GPU, explicitly select the CUDA build in the same PowerShell window before installing:

$env:EDGESPEAK_WINDOWS_ACCELERATOR = "cuda"
irm https://edgespeak.com/install.ps1 | iex

You can run edgespeak-cli login as soon as installation finishes. If another open terminal cannot find the command, reopen it so the PATH update takes effect.

Activate

Use edgespeak-cli login to sign in through your browser: new accounts start a 7-day trial and purchased accounts activate this device. For an instant device-bound trial with no browser or account, use edgespeak-cli trial. Already have a license key? Use edgespeak-cli activate <KEY>, or --stdin to keep it out of shell history.

edgespeak-cli login

edgespeak-cli trial            # instant device trial, no account or browser
edgespeak-cli activate <KEY>
edgespeak-cli activate --stdin   # paste the key, then Ctrl-D

Workflows

Transcribe with optional speaker labels, align known text, segment transcripts, synthesize speech, manage voices, inspect models, or open the Realtime demo. Most one-shot commands start the bundled local engine when the app is not running; serve instead keeps a headless HTTP service running for repeated API calls.

edgespeak-cli transcribe meeting.m4a -o meeting.json   # word-level timing in JSON
edgespeak-cli transcribe meeting.m4a --diarize -o speakers.json
edgespeak-cli transcribe meeting.m4a -o transcript.srt
edgespeak-cli align meeting.m4a --text-file transcript.txt -o aligned.json
edgespeak-cli segment --file transcript.txt -o sentences.json
edgespeak-cli speech "Read this aloud" -o speech.wav
edgespeak-cli voices list
edgespeak-cli models list
edgespeak-cli status
edgespeak-cli demo
edgespeak-cli update

Headless API service

Run an OpenAI API-compatible HTTP service without the desktop app. By default, it listens only on http://127.0.0.1:1118/v1. To let another trusted machine call it, set EDGESPEAK_API_KEY on the service host (the machine running the service) and explicitly enable a non-loopback listener:

export EDGESPEAK_API_KEY="your-long-random-secret"
edgespeak-cli serve --host 0.0.0.0 --port 1118 --allow-remote

Set the language-model context window when the service starts with --context-tokens. It defaults to 8192 tokens and applies to local language-model workers started by that process. Headless Server does not read context choices saved in the desktop app.

edgespeak-cli serve --context-tokens 32768

For native video requests, frame sampling is selected automatically from the model catalog—2 FPS for Qwen 3.5/3.6 and 1 FPS for Gemma 4—so there is no video-FPS flag to set. Context size and sampled video frames share the same memory and token budget. See Native video input and context window for the request shape, dependencies, and resource impact.

The client (the machine making the request) must use the service host's actual LAN or VPN address, such as http://192.168.1.50:1118/v1, never 0.0.0.0. Model loading and inference run on the service host. For a client example, firewall configuration, browser origins, SSH tunnels, and HTTPS security boundaries, read Call the API from another machine.

Agents

Drive it from agents through the EdgeSpeak Skill, or wire the local gateway and MCP into automation and CI.

Privacy

Speech workflows and installed local models run on this device. If you explicitly select a model reported as execution_location=remote, only that request is sent to its configured provider.

Generate with a local model

Start the desktop app, download and load a model, then use Responses, Chat Completions, the CLI, or MCP. Installed local models keep prompts and images on this device.

edgespeak-cli models load Qwen/Qwen3.5-4B
edgespeak-cli generate "Summarize the meeting" --model Qwen/Qwen3.5-4B
edgespeak-cli generate "Describe this image" --image screenshot.png --json