Show HN: Deepgram releases Deepgram CLI (`dg`) an agent-aware CLI (cli.deepgram.com)

Deepgram has launched the Deepgram CLI (`dg`), a command-line interface designed to streamline speech AI functionalities directly in the terminal. This tool enables developers to easily transcribe, synthesize, and stream audio with just one command. Key features include advanced speech-to-text capabilities, such as speaker diarization, smart formatting, topic detection, and summarization. Additionally, users can enjoy live streaming with real-time transcripts and word-level timing, as well as text-to-speech generation using Aura voices. The significance of this release lies in its focus on enhancing productivity for developers in AI and machine learning by integrating powerful speech processing features directly into a familiar terminal environment. The CLI is agent-aware by default, meaning it can automatically recognize and utilize various AI models like Claude Code, Aider, and Codex, simplifying the workflow without requiring complex setups. This innovation not only expedites audio processing tasks but also broadens accessibility for developers looking to leverage speech AI in their applications.