prompting guide

silk muga 1#

write your prompt in latin script only — e.g. yeh ek test message hai, not यह एक टेस्ट मैसेज है.

silk muga 1 is a hinglish emotion-TTS model with two control surfaces: a paragraph tone set by the selector or a [tone] marker, and discrete inline events (laugh, chuckle, sigh) you place wherever you want them to sound.

1. paragraph tone#

the voice tone selector sets the default tone for every paragraph. type [ at the start of a paragraph to insert a specific [tone] marker for that paragraph.

rules

• one tone marker per paragraph; a blank line starts a new paragraph.
• paragraphs without an explicit marker use the selected default tone; [neutral] is the default when nothing else is selected.
• the marker applies to the whole paragraph, even if it is inserted after the first word.

2. inline events#

type these directly in the transcript at the position you want the sound. three are supported:

rules

• lowercase, angle brackets, no spaces inside. <laugh>, never <Laugh> or < laugh >.
• space on both sides when between words. never mid-word.
• position-sensitive. <laugh> kya baat hai sounds different from kya baat hai <laugh>.
• stack at most two. <laugh> <laugh> for a longer/harder laugh. three or more becomes unstable.

3. tone–event compatibility#

events have to match the tone. laughter belongs to high-energy positive states; sighs belong to low-energy reflective ones. mix them — a laugh in a sad line, a sigh in an excited shout — and the model fights itself, because the training set has almost no examples of those combinations.

✓✓ best · ✓ ok · ~ rare · ✗ avoid

4. good vs bad#

[sad] <laugh> sab kuch khatam ho gaya
[neutral] <laugh> aaj ka mausam saaf rahega
[angry] <chuckle> tumne phir galti ki
[happy] <sigh> kya mast din tha aaj
[excited] <sigh> jeet gaye!
[whisper] <laugh> sab so rahe hain

[happy] <laugh> Yaar tumne phir wahi joke maara!
[excited] <laugh> Bhai jeet gaye, vishwas nahi ho raha!
[sad] <sigh> Pata nahi yaar, kuch samajh nahi aata.
[whisper] <sigh> Itna lamba din tha, thak gayi hoon.

5. language register#

training data is hinglish — romanised hindi with english code-mixing. the model speaks that best.

avoid

• devanagari. the model saw zero hindi script. मैं ठीक हूँ produces garbage.
• other indian languages (tamil, bengali, marathi, bhojpuri).
• heavy regional dialects (very bambaiyya, very punjabi).

6. length#

silk muga 1 is built around 2 to 30 second utterances and extrapolates reliably up to around 40s. beyond that, tone drifts, pacing slips, and you start seeing repetitions or cutoffs.

• 2 to 30s — sweet spot, one to three sentences.
• ~30 to 40s — still works for longer monologues.
• beyond 40s — split across prompts.

7. examples#

silk mulberry 1.5#

write the voice in natural language. the model picks up on attributes you mention. the lists below are the vocabulary it understands; weave them into a single sentence rather than listing them as fields.

1. inline tags#

drop these tags anywhere in the text to trigger a sound. they render as part of the performance, not as words.

<laugh>  <laugh_harder>  <sigh>  <chuckle>  <gasp>  <angry>  <excited>
<whisper>  <cry>  <scream>  <sing>  <snort>  <exhale>  <gulp>  <giggle>
<sarcastic>  <curious>

2. voice attributes#

mention any of these in your description.

• age — 20s, 30s, 40s
• accent (global) — american, british, middle_eastern, asian_american, indian
• accent (indian regional) — hindi, punjabi, bihari, south_indian, bengali, rajasthani, marathi, gujarati, kashmiri, assamese, odia, telugu, kannada, malayali, haryanvi, chhattisgarhi
• pitch — low, normal, high
• timbre (realistic) — deep, warm, gravelly, smooth, raspy, nasally, throaty, harsh, whisper
• timbre (creative) — adds robotic, ethereal to the realistic set
• pacing — very slow, slow, conversational, brisk, fast, very_fast
• emotion — neutral, energetic, excited, sad, sarcastic, dry, crying, angry
• intensity — low, med, high
• register — formal, neutral, casual

3. speaking role#

pick a role from a domain to anchor the delivery style.

• social — youtube_vlogger, social_media_creator, influencer_voice, streamer_companion
• podcast — podcast_host, interviewer
• commercial — ad_narrator, brand_spokesperson, product_demo_voice, sales_pitch_voice
• education — elearning_instructor, kids_story_voice
• support — customer_support_agent, virtual_receptionist, healthcare_assistant
• entertainment — storyteller, social_media_reaction, meme_voice
• corporate — explainer_video_voice, event_host, corporate_training_narrator
• viral — short_form_narrator, meme_voice

4. creative-only attributes#

available when you want a non-realistic timbre (e.g. characters, stylized voices).

animated_cartoon  ai_machine_voice  alien_scifi  seductively  flirty  anime
cyborg  pirate  dark_villain  demon  gangster  mafia  dramatic_narrator
mythical_godlike_magical  spy  vampire  alpha

5. examples#

description: a warm 30s hindi accent voice, conversational pacing, casual
register, sounds like a podcast host walking you through a story.

transcript: aaj ka episode thoda alag hai. <chuckle> ek minute ke liye seedha
baith jao.

description: a high pitched 20s american voice, excited, very fast pacing, like
a streamer reacting live.

transcript: oh my god did you see that play, that was insane.

description: a deep gravelly low pitched 40s british voice, slow pacing, formal
register, dramatic narrator.

transcript: the door creaked open. nobody was there. and yet, something watched.