TikTok Adds Voice Comments, Polls, and Multi-Photo Replies to Its Comment Sections
TikTok announced four new interactive comment features on September 3, 2026: voice comments, comment polls, photo carousel comments, and Live Photo comments. The additions expand the comment section beyond the text, photo, and video sticker comments the platform already supported. TechCrunch
Voice Comments let users record an audio message of up to one minute by tapping a microphone icon in the comment box, replacing typed text with recorded speech. Anyone viewing the thread can tap to listen. The feature requires no typing and is rolling out globally over the month following the announcement, restricted to users aged 18 and older. According to a TikTok spokesperson, voice comments are subject to the same Community Guidelines as other platform content and are monitored using both human and automated detection tools, including speech-to-text transcription — the process of converting spoken audio into written text that a moderation system can then scan. TechCrunch
Comment polls, previously available to a limited set of users, are now globally available to creators on their own videos. Creators can set up to five voting options and define a timeframe for the poll, with results displayed directly in the comment section. TechCrunch
Photo carousel comments raise the per-comment image limit from one photo to nine, and are also rolling out globally over the month following the announcement. Live Photo comments, already available globally, let users upload Apple Live Photos that display a brief burst of motion as a preview before the full image is viewed. TechCrunch
TikTok reports that more than 1.7 billion comments are shared across its platform every day, according to the company's newsroom. The existing comment infrastructure already supported text, photo, and video sticker comments before these additions.
The technically notable detail is the speech-to-text transcription layer applied to voice comments. Audio content moderation at scale has long been harder than text or image moderation, and transcribing voice comments into text before applying automated policy enforcement is the same architectural pattern platforms use for voice messages in messaging products. The approach lets TikTok reuse its existing text-based detection pipelines rather than building a parallel audio-classification system from scratch. It also creates a structural tension: transcription accuracy varies across languages and accents, which means enforcement consistency will depend on the quality of TikTok's speech-to-text models across its global user base.
The 18-and-over restriction on voice comments signals that TikTok is treating audio as a higher-risk medium than text or images, at least for initial rollout. This aligns with a pattern across social platforms where voice and live audio features have historically drawn faster regulatory and safety scrutiny than text-based equivalents.
The broader trajectory here is that TikTok's comment section is converging on the expressiveness of its main feed. Comments already supported photo and video stickers. Adding voice, polls, multi-photo carousels, and Live Photos means a comment can now carry nearly the same media richness as a post. At 1.7 billion comments per day, even a modest shift in engagement per comment translates into substantial additional content volume flowing through TikTok's moderation infrastructure.
For developers and content tools vendors watching TikTok's creator ecosystem, the comment polls feature is the one most likely to generate structured engagement data. Polls with up to five options and time-bound windows produce discrete, queryable interaction data that sits somewhere between a like and a full comment in terms of user effort. That structured signal could feed recommendation ranking models in ways that free-text comments cannot.
The rollout timeline is staggered. Comment polls and Live Photo comments are already globally available. Voice comments and photo carousel comments are rolling out over the month following the September 3 announcement.
