Oliver Wolfson | Technical Product Builder

Here's a comprehensive breakdown of the audio tags that work with ElevenLabs Eleven v3. ElevenLabs v3 introduces emotional control through audio tags — you can direct voices to laugh, whisper, act sarcastic, or express curiosity, among many other styles. Note that the voice you choose and its training samples will affect tag effectiveness — some tags work well with certain voices while others may not.

🎭 Emotion Tags

Tag	Effect
`[happy]`	Upbeat, cheerful delivery
`[sad]`	Somber, downcast tone
`[angry]`	Forceful, irritated delivery
`[excited]`	High energy, enthusiastic
`[fearful]`	Nervous, scared tone
`[disgusted]`	Repulsed or distasteful delivery
`[surprised]`	Shocked or startled reaction
`[curious]`	Inquisitive, wondering tone
`[confused]`	Uncertain, puzzled delivery
`[bored]`	Flat, disengaged tone
`[nervous]`	Anxious, hesitant delivery
`[disappointed]`	Deflated, let-down tone
`[proud]`	Confident, self-assured delivery
`[relieved]`	Exhaled, tension-releasing tone
`[hopeful]`	Optimistic, forward-looking delivery
`[melancholic]`	Wistful, reflective sadness
`[panicked]`	Rushed, frantic delivery
`[calm]`	Measured, serene tone

🗣️ Delivery / Style Tags

Tag	Effect
`[whispers]`	Soft, hushed voice
`[shouting]`	Loud, projected voice
`[sarcastic]`	Ironic, mocking delivery
`[serious]`	Formal, no-nonsense tone
`[sympathetic]`	Warm, empathetic delivery
`[dramatic]`	Theatrical, heightened emotion
`[monotone]`	Flat, emotionless delivery
`[cheerful]`	Light and pleasant
`[stern]`	Firm and authoritative
`[gentle]`	Soft and kind
`[condescending]`	Patronizing, talking-down tone
`[matter-of-fact]`	Direct, neutral delivery
`[encouraging]`	Motivating, supportive tone
`[conspiratorial]`	Secretive, leaning-in tone
`[pleading]`	Begging, desperate delivery

😂 Non-Verbal / Vocal Sounds

Tag	Effect
`[laughs]`	Natural laugh
`[chuckles]`	Soft, short laugh
`[giggles]`	Light, playful laugh
`[sighs]`	Exhale of resignation or relief
`[gasps]`	Sharp inhale of surprise
`[crying]`	Teary, emotional voice
`[sobbing]`	Heavy, uncontrolled crying
`[yawns]`	Tired, sleepy exhale
`[clears throat]`	Throat-clearing sound
`[sniffles]`	Nasal, holding-back-tears sound
`[groans]`	Low, pained or annoyed sound
`[screams]`	High-intensity vocal outburst

🔊 Sound Effect Tags

Sound effect tags can be inserted inline to trigger environmental or action sounds without being spoken aloud.

Tag	Effect
`[applause]`	Crowd clapping
`[gunshot]`	Firearm sound
`[explosion]`	Blast/explosion sound
`[door slam]`	Door closing hard
`[thunder]`	Thunderclap
`[phone ringing]`	Phone sound
`[glass breaking]`	Shattering glass
`[footsteps]`	Walking sounds

⚡ Pacing / Intensity Tags

Tag	Effect
`[slowly]`	Drawn-out, deliberate pacing
`[quickly]`	Faster, urgent delivery
`[urgently]`	Pressured, time-sensitive tone
`[hesitatingly]`	Uncertain, stop-start delivery
`[breathlessly]`	Out-of-breath, rushing tone

📌 Important Tips

Stability setting matters. For maximum expressiveness with audio tags, use the Creative or Natural stability setting. The Robust setting reduces responsiveness to directional prompts.
Voice compatibility is key. If a voice is naturally loud and you apply [whispering], it likely won't work well — the base voice needs to be compatible with the tag.
Tags are not spoken aloud — they act as invisible stage directions.
Combine tags for layered effects, e.g.: [whispers][nervously] I think someone is watching us.
v3 does not support SSML <break> tags — use ellipses … or dashes — for pauses instead.
The tag system is flexible and not strictly limited to this list — v3 can interpret many natural-language descriptors in brackets, so experimenting with your own is encouraged.