feat(tts/gemini): Added Gemini TTS support

[codejutsu1]

This PR adds Gemini TTS (Text-to-Speech) support. Fixes #588

Notable Features

• Audio outputs in raw audio format .pcm - users will need to manually convert using ffmpeg to wav or mp3
• Multi-speaker configuration support

Usage Example

$response = Prism::audio()
    ->using(Provider::Gemini, 'gemini-2.5-flash-preview-tts')
    ->withInput('TTS the following conversation between Joe and Jane:
        Joe: How\'s it going today Jane?
        Jane: Not too bad, how about you?')
    ->withVoice('Enceladus') 
    ->withProviderOptions([
        'multiSpeaker' => [
            [
                'speaker' => 'Joe',
                'voiceName' => 'Kore',
            ],
            [
                'speaker' => 'Jane',
                'voiceName' => 'Puck',
            ],
        ],
    ])
    ->asAudio();

[cmayorgahilario]

up