speechall-cli

What This Skill Does

The speechall-cli skill integrates the powerful Speechall API directly into your terminal environment. It acts as an interface for high-accuracy speech-to-text transcription, supporting a vast array of providers like OpenAI, Deepgram, and AssemblyAI. Whether you are dealing with raw audio files or complex video formats, this tool handles the heavy lifting of extraction and transcription. It is designed for users who require command-line efficiency to handle large volumes of media, offering features like speaker diarization, custom vocabulary boosting, and flexible output formats such as SRT and VTT for subtitle creation.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/atacan/speechall-cli

For the underlying CLI tool itself, ensure you have Homebrew installed on macOS or Linux, then run: brew install Speechall/tap/speechall If you prefer not to use Homebrew, you can manually download the latest binary from the official GitHub repository and move it into your system's PATH. After installation, verify the setup by running speechall --version.

Use Cases

• Professional Transcription: Automatically convert long interview recordings or meeting audio into structured text documents.
• Content Creation: Generate subtitle files (SRT/VTT) from video presentations for accessibility or social media captions.
• Domain-Specific Accuracy: Improve recognition of technical terminology in medical, legal, or engineering audio using the custom vocabulary feature.
• Speaker Analysis: Use diarization to automatically attribute dialogue to specific speakers in multi-participant recordings.

Example Prompts

1. "Transcribe this interview file named recording.mp3 and save the output as an SRT file with speaker diarization enabled for two people."
2. "Can you list all the available STT models supported by speechall so I can see which ones support high-accuracy German transcription?"
3. "Transcribe my meeting video titled sync.mp4, use the Deepgram Nova-2 model, and boost the vocabulary for the term 'OpenClaw'."

Tips & Limitations

To avoid passing your API key in plain text within commands, always set the SPEECHALL_API_KEY environment variable in your shell configuration. When using diarization, performance improves significantly if you provide an estimate of the number of speakers expected in the audio file. Note that while the tool handles video files by extracting audio, very large files may require significant processing time; for long-form content, consider splitting files beforehand. Always ensure you have sufficient credits on your Speechall account for the models you intend to use.