Short answer: No—those inline tags aren’t supported for ElevenLabs v3.
[speed 1.2] is not a recognized v3 audio tag. In v3, speaking rate is controlled via the speed setting (not inline text). Set speed to 0.7–1.2 (default 1.0) in Voice Settings or via the API (voice settings speed field), e.g., using the “Edit voice settings” endpoint. (elevenlabs.io)
[neutral] is not a documented audio tag for v3. “Neutral” appears in docs as a voice characteristic or stability guidance, not as a tag. The documented v3 tags cover emotions/delivery (e.g., [happy], [sad], [whispers]) and non‑verbal cues. (elevenlabs.io)
For pauses in v3, use [pause], [short pause], or [long pause]. SSML <break> works on other models but not on Eleven V3. (help.elevenlabs.io)
References:
Best practices (pace and speed range 0.7–1.2) and v3 prompting/tag guidance. (elevenlabs.io)
odd sapphire
daring chasm