You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
docs: improve --help for text chat, search, speech synthesize
- text chat --model: list available models (M2.7, M2.7-highspeed)
- text chat --message: clarify role: prefix usage with examples
- text chat --messages-file: show expected JSON format
- text chat --temperature: explain 0.0-1.0 range and default (0.7)
- text chat --top-p: explain probability mass concept and default (0.95)
- text chat --tool: mention OpenAI compatibility
- speech synthesize --model: list all models with characteristics
- speech synthesize --text: note 10,000 char limit and SSML support
- speech synthesize --voice: point users to 'mmx speech voices' command
- speech synthesize --speed/volume/pitch: add ranges and defaults
- speech synthesize --format: list all formats (mp3, wav, ogg, flac)
- speech synthesize --sample-rate: list common values
- speech synthesize --language: list supported language codes
- speech synthesize --subtitles: explain use case (karaoke/video sync)
- speech synthesize --pronunciation: clarify word/pronunciation format
- speech synthesize --stream: add pipe-to-player example
- search query: clarify output includes titles/URLs/snippets
- search query: add pipeline example (search → text chat)
{flag: '--language <code>',description: 'Language code to boost pronunciation accuracy (e.g. "en", "zh", "ja", "ko", "es", "fr", "de"). Auto-detected from text if not specified.'},
30
+
{flag: '--subtitles',description: 'Include word-level subtitle timing data in the response. Useful for karaoke or video syncing.'},
{flag: '--message <text>',description: 'Message text (repeatable). Prefix with role: to set role — e.g. "user:Hello", "assistant:Hi!", "system:You are helpful"',required: true,type: 'array'},
85
+
{flag: '--messages-file <path>',description: 'JSON file with messages array (use - for stdin). Format: [{"role": "user", "content": "..."}]'},
86
+
{flag: '--system <text>',description: 'System prompt to set assistant behavior'},
87
87
{flag: '--max-tokens <n>',description: 'Maximum tokens to generate (default: 4096)',type: 'number'},
88
-
{flag: '--temperature <n>',description: 'Sampling temperature (0.0, 1.0]',type: 'number'},
0 commit comments