Smart TTS with real content filtering, live event alerts, beautiful OBS overlays, and Spotify integration — everything a TikTok streamer needs in one free app.
Everything you need to run a polished TikTok LIVE, in one lightweight desktop app.
Follow, gift, share, like and chat events arrive the moment they happen. A live dashboard keeps everything visible at a glance so you never miss a moment.
Reads chat and events aloud using Google's high-quality TTS voices. Hard-blocks slurs, asterisks profanity, detects spam, enforces message length limits, and supports 16 languages — all built in.
Six browser-source overlays ready for OBS, Streamlabs or TikTok Studio: live chat dock, follow alerts, gift alerts, share alerts, like counter, and a Spotify now-playing widget.
Connect Spotify for the full experience. Viewers can type !play artist - song in chat to queue tracks directly. Plus a live OBS overlay showing album art, track name and a real-time progress bar.
Play custom MP3 files for follows, gifts, shares and likes. Bundled defaults get you started instantly — or drop in your own audio.
Generic TTS tools read everything — spam, slurs, copy-pasted junk, usernames you've already banned. TikTalk was built differently.
You get complete, per-rule control over what gets read aloud. Your stream sounds professional. Your chat stays welcoming. You stay in control.
Every TTS message passes through an ordered pipeline. A single hit at any layer silently drops or cleans the message — your viewers only ever hear what's safe.
Choose who can trigger TTS — everyone, followers only, subscribers, or a custom allowlist. Anonymous accounts and one-time trolls never get a voice.
Configurable minimum and maximum character limits. Short emoji-only noise and wall-of-text pastes are both skipped before any expensive processing starts.
Maps 250+ lookalike characters — full Cyrillic, Armenian, Greek and fullwidth Latin blocks — to their ASCII equivalents so every subsequent check works on clean text.
50+ slur and hate-speech root forms, each compiled with suffix tolerance — one entry automatically covers plurals, verb forms and common extensions without extra rules.
Double Metaphone algorithm flags words that look different but sound identical when spoken. Creative misspellings designed to slip past word lists are caught here instead.
Regex patterns for self-harm triggers (kys, kill yourself, top yourself), Nazi phrases (sieg heil), and TTS phonetic attack strings that sound like slurs when read aloud.
Detects character floods (aaaaaaaa…), three-or-more repeated words, and ALL-CAPS outbursts above a threshold. No more spam bots monopolising your TTS queue.
Softer swear words are replaced with asterisks rather than dropped entirely. The message is still read aloud — just cleaned up — keeping your stream welcoming without silencing everything.
Every feature was built from scratch, from the ground up. Here's the honest timeline of what's shipped, what's being tested, and what's next.
Multi-process Electron app with IPC bridging, secure preload sandboxing, and a clean main/renderer split established from day one.
WebSocket connection to TikTok's live API with automatic reconnect, latency monitoring and event deduplication across all event types.
Real-time event feed with tabbed views, live counters, and colour-coded event types — all updating instantly without any page reload.
High-quality cloud TTS with per-language voice selection, playback rate control, and a priority queue to prevent audio stacking during busy chat.
Full gift event pipeline — gift name, quantity, per-gift diamond value, running stream totals, and gifter leaderboard ranking built in.
Full Spotify auth flow with automatic token refresh, current track polling, and WebSocket push to the live Now Playing overlay.
!play, !skip, !queue and !np commands parsed from live chat in real time — lets your audience DJ your stream without touching the app.
MP3 playback for follows, gifts, shares and likes — bundled defaults included, with drop-in support for your own audio files per event type.
Chat dock, follow/gift/share alert banners, animated likes counter and Spotify Now Playing widget — each on a dedicated URL, zero config needed.
250+ lookalike character mappings covering Cyrillic, Armenian, Greek and fullwidth Latin — closes the most common filter-bypass vector used in TikTok chat.
Algorithmically flags words that sound like blocked terms even when creatively misspelled — the layer most TTS tools skip entirely.
Access gate → length check → normalisation → slur block → phonetic detection → harm phrases → spam detection → profanity asterisking — every message, every time.
Role-based TTS access (everyone / followers / subscribers / custom allowlist), per-user muting, and independent enable/disable per event type.
Every setting survives app restarts — TTS toggles, filter thresholds, Spotify credentials, overlay server port, and sound effect paths all saved automatically.
NSIS-based installer with desktop shortcut, Start Menu entry, and electron-updater wired for silent background updates on app launch.
Load testing the full filter pipeline and audio queue under high-volume chat conditions, catching edge cases before any public-facing release.
Real streamers running TikTalk on real streams, surfacing UX friction, unexpected filter behaviour and performance issues before v1.0 ships.
Your images, your format, your brand. Full control over how each overlay looks — custom backgrounds, colours, fonts and layout, built around what you actually want on screen.
Full public launch once BETA feedback is resolved. Free for all Windows users, no account or subscription required, ever.
Download TikTalk and go live with smart TTS, beautiful overlays and real-time event alerts — all for free.
🔒 In Beta TestingTikTalk is free and always will be. Donations are never expected or necessary — but if the app has made your streams better, buying a coffee helps keep the project going.
Buy me a coffee