Is guiding the pronunciation possible?

[S-T-K]

Is there a way to influence the pronunciation, pacing, and emotion in the TTS output?
For instance, in ElevenLabs, placing quotation marks around a word can create stronger emphasis. The only methods I’ve found to actively control pacing involve using punctuation marks (e.g., . , ; : ? !) or adding ellipses or dashes for pauses, see https://github.com/erew123/alltalk_tts?tab=readme-ov-file#-tricks-to-get-the-model-to-say-things-correctly
Any other adjustments appear to be ignored.

[finefin]

That depends on the model. A new model was released a few days ago that seems to be able to do what you ask for:
https://huggingface.co/OuteAI/OuteTTS-0.3-1B