Kling TTS Emotional Speech Generation

Problem

When an agent needs to turn prompts, subtitles, or copy into playable audio, calling a cloud TTS SDK directly brings in key setup, async task polling, and error handling. keling tts wraps this into a single dlazy keling-tts command: provide text and parameters, then receive Kling TTS speech output or task status, making emotional speech generation easier to embed in local workflows.

How It Works

The skill is a thin client for the dLazy hosted API. On invocation, text and parameters are sent to api.dlazy.com for inference; if local file paths are supplied, the CLI uploads them to dLazy media storage so the model can read them. Generated assets are typically returned as URLs on files.dlazy.com.

• Authentication: uses a dLazy API key, or DLAZY_API_KEY for a single invocation.
• Synchronous output: waits by default and returns the finished speech result.
• Async mode: with --no-wait, returns task.generateId and status for later polling.
• Parameter references: supports -, @N, @N.path, and @stdin for piping upstream values.
• Error signals: 401 means the key is missing or invalid; 503 may indicate insufficient balance or service errors.

Boundaries

This is best for converting short text, scripts, or generated copy into Kling TTS speech. It is not a good fit for fully offline inference, self-hosted deployment, or avoiding dLazy account billing. If the result reports insufficient_balance or unauthorized, check the dLazy account credits or API key first.