Play.ht had the largest voice library on the market: 900+ voices and 142+ languages. They achieved the catalog by aggregating across multiple synthesis engines (their own PlayHT 2.0 plus Google Cloud, AWS Polly, and third-party providers), which meant voice quality varied. FreeTTS uses one engine for 322 voices and 75 languages with consistent quality. The choice used to be: Play.ht for catalog size and WordPress integration, FreeTTS for cleaner consistency and a free public API. Same dollars, different shapes — preserved below exactly as the matchup stood before the shutdown.
None of this is available anymore — Play.ht shut down on December 31, 2025. Need these capabilities today? See our Play.ht alternative guide.
All 50+ rows, grouped into 10 categories. FreeTTS pricing is current; Play.ht figures are the last prices the company published before its December 31, 2025 shutdown, preserved as an archival record.
| Feature | FreeTTS | Play.ht |
|---|---|---|
| Free tier | ||
| Free chars per month | 60,000 | Trial only (~12,500 chars) |
| Audio watermark on free | Short audio tag at end | Trial restrictions |
| Commercial use on free tier | ||
| Voices available on free | All 322 neural voices | Limited subset |
| API on free | Yes, 15 req/min, no key | |
| Paid pricing | ||
| Entry tier | PRO $19/mo | Creator ~$31/mo |
| Mid tier | Creator $39/mo | Pro ~$49/mo |
| Enterprise tier | (custom on request) | Enterprise custom |
| Lifetime option | $199 PRO / $349 Creator | |
| Annual discount | 20% off | ~17% off |
| Refund window | 7 days | Per their TOS |
| Volume per dollar | ||
| Chars on entry paid tier | 1,000,000 | Unlimited words on standard voices |
| Chars on mid tier | 5,000,000 | Premium voice volume gated |
| Cost at entry tier | $19/mo | $31/mo (or $49 for premium voices) |
| Cost per 1,000 chars at mid tier | $0.0078 | Variable, ~$0.01-0.05 |
| Voices and languages (where Play.ht won on catalog) | ||
| Total voices . Play.ht aggregated across PlayHT 2.0, Google Cloud, AWS Polly, and third-party providers engines | 322 (marketed as 400+) | 900+ (multi-engine aggregate) |
| Voice quality consistency . Voice character drifted across engines on Play.ht | Single-engine, consistent | Multi-engine, varied |
| Languages (unique) | 75 | 142+ |
| Locale variants | 142 | Many per language |
| HD flagship voices | 12 verified HD multilingual* | PlayHT 2.0 voices |
| Female / Male balance | 163 / 159 | Mixed across engines |
| Community voice library | Limited | |
| Quality and expressiveness | ||
| Expressive speaking styles | 95 (with intensity slider) | Limited per voice |
| Voice cloning | Yes (PRO 3 / Creator 10) | Yes, on higher tiers |
| Min sample for cloning | 30 seconds | 30 seconds (instant) |
| Cloning quality (subjective) | Good for most use cases | Varies by source engine |
| Multi-voice dialogue | Yes, in /studio | |
| Pronunciation library per project | ||
| Output and formats | ||
| MP3 export | ||
| WAV export | PRO+ | |
| OGG export | PRO+ | |
| FLAC export | ||
| SRT subtitle generation | Native, every gen | Manual / post-processing |
| Multi-voice per file | PRO 2 / Creator unlimited | |
| SSML support | Full W3C SSML | Engine-dependent |
| IPA phoneme tags | PlayHT 2.0 voices only | |
| Workflow integrations (where Play.ht won) | ||
| WordPress plugin . Auto-converted blog posts to audio with embedded player | ||
| Podcast hosting integration . Published generated audio to RSS feeds (Apple Podcasts, Spotify) | ||
| Zapier / make.com | ||
| PDF to audiobook tool | Yes, with chapter detection | |
| Multi-speaker dialogue studio | Yes, in /studio | |
| Developer / API | ||
| Free public API | Yes, no key, 15 req/min | |
| API rate limit on PRO | 200 req/min | Varies by plan |
| API rate limit on top tier | 1,000 req/min (Creator) | Higher on Enterprise |
| Streaming endpoint | ||
| Webhooks | ||
| Browser surface | ||
| Chrome extension (free) | v1.3 with MP3 download | |
| Mobile app . Both were web-only; FreeTTS is fully mobile-responsive | ||
| Right-click MP3 download from any page | ||
| Voice picker in popup | All 322 voices | |
| Trust and policy | ||
| Privacy: voice clone samples retained? | No, deleted after use | Per their TOS |
| Voice clone samples used for training? | ||
| GDPR / EU hosting | Yes, German data center | Mixed (US-primary) |
| SOC 2 / enterprise compliance | Yes, on Enterprise tier | |
| Public roadmap | Limited | |
*HD multilingual voices: 12 verified in the public API today (Andrew, Ava, Brian, Emma, William, Vivienne, Remy, Seraphina, Florian, Giuseppe, Hyunsu, Thalita). Marketed as "30 HD voices" when counted with their per-voice expressive style variants. Play.ht's 900+ count aggregated across PlayHT 2.0, Google Cloud, AWS Polly, and third-party providers engines (archival figure — the service shut down December 31, 2025).
A narration-heavy minute uses ~1,300 characters. Use that to estimate your monthly volume. Play.ht costs shown are its final published pricing (the service shut down December 31, 2025).
Per-month math at entry tier (archival): FreeTTS PRO $19 vs Play.ht Creator's final $31. Annual saving on FreeTTS PRO was $144/year vs Play.ht Creator. Over 2 years, that is $288 saved, more than enough to fund a Lifetime upgrade ($199) with money to spare.
If you are coming from Play.ht, here is the closest FreeTTS HD multilingual voice for each of their former flagship PlayHT 2.0 voices. PRO HD voices, not free-tier voices.
Both natural, conversational warmth. Jennifer is slightly more expressive on long-form; Ava holds tone better across languages.
Closest pairing for explainer and corporate content. Brian is neutral and even.
Both work for documentary and YouTube narration. Andrew is documentary-leaning; Mark slightly friendlier.
Both upbeat, suitable for marketing scripts and YouTube intros.
Hear all 322 FreeTTS voices Play.ht's voice library went offline with the Dec 31, 2025 shutdown
Play.ht was bigger and more bundled, and that held right up to the December 31, 2025 shutdown. We are not going to pretend it was not.
If catalog size was your primary metric, Play.ht won. They aggregated voices across their PlayHT 2.0 engine plus Google Cloud, AWS Polly, and third-party providers engines, totaling 900+ voices. Quality varied between engines. FreeTTS has 322 verified voices on a single engine. For 'I want maximum voice variety in one tool' use cases, Play.ht was the bigger library.
Play.ht had the broadest language coverage on the market, including some indigenous and less-served languages that FreeTTS does not have. If you published content in exotic languages outside the top 75, Play.ht was meaningfully wider.
Their WordPress plugin auto-converted blog posts to audio and embedded a player on the post. Best-in-class "add audio to your blog" workflow while it operated. FreeTTS has no WordPress plugin. For bloggers wanting to ship audio versions of every post automatically, Play.ht was structurally better suited.
You could generate audio in Play.ht and publish it directly to an RSS feed for Apple Podcasts and Spotify, no separate podcast host required. FreeTTS generates the MP3 and you bring your own host (Buzzsprout, Anchor, Spotify for Podcasters). Less bundled.
For live voice agents, real-time dubbing, and voice chatbots, Play.ht had a streaming endpoint that shipped audio chunks as they generated. FreeTTS is currently batch-only. For live applications, this was a real reason to use Play.ht.
One Play.ht subscription gave you access to voices on multiple synthesis engines. If you wanted to compare Google Cloud vs another provider vs PlayHT 2.0 quality on the same script without managing four accounts, Play.ht was convenient. FreeTTS is single-engine by design.
PRO at $19/month vs Play.ht Creator's final $31/month. $12/mo saved equals $144/year, $288 over two years. For most YouTube and podcast creators that is enough to fund a Lifetime upgrade ($199 one-time).
Single neural engine for all 322 voices. Voice character, pacing, and tone are uniform across the library. Play.ht aggregated engines so a voice powered by a third-party engine sounded different from a voice powered by Google in tonality and rhythm. For brand consistency across episodes, FreeTTS is the cleaner pick.
FreeTTS is focused: paste text, pick voice, generate, download. Play.ht was denser by design (multi-voice editor, podcast hosting, WordPress, multi-engine voice browser) which added onboarding friction. For first-time users producing a single MP3, FreeTTS is faster to a finished file.
15 requests per minute on the free tier, 200 requests per minute on PRO, 1,000 requests per minute on Creator. No API key required for free tier. Play.ht had API access only on paid plans, with key authentication required.
FreeTTS free tier is real: 60,000 chars/month, downloadable MP3 with short audio tag. Play.ht had a trial that ended and converted you to paid. For creators testing a few free generations before committing, FreeTTS lets you ship without a credit card.
$199 one-time for PRO Lifetime. $349 one-time for Creator Lifetime. Limited to the first 100 buyers, real countdown on the site. Play.ht never had a lifetime offering. Two years of Play.ht Creator at $31/mo would have been $744; FreeTTS PRO Lifetime saves $545 minimum against that.
Every FreeTTS generation produces an MP3 plus a perfectly synced SRT subtitle file. Word-level timing comes directly from the synthesis engine. Play.ht did not generate SRT directly; you built them yourself or ran a separate tool.
FreeTTS includes a free PDF-to-audiobook tool that auto-detects chapters and generates per-chapter MP3s. Play.ht had no equivalent product. For audiobook producers, this was a real workflow gap.
Both tools were web-first, but how did they handle "I am reading an article and want to hear it"?
| Feature | FreeTTS extension v1.3 | Play.ht |
|---|---|---|
| Free, no signup | ||
| Right-click any text on any page | ||
| MP3 download from popup | ||
| Voice picker in popup (322 voices) | ||
| Manifest V3 | ||
| Open source listing |
Real timing. Five steps. No drama.
Play.ht shut down on December 31, 2025 — subscriptions ended with the platform and no one has been billed since. There is nothing to cancel; recover any scripts or reference audio from your local backups, then move to step 2.
Visit freetts.org. The first three generations work without signup. After that, free account requires only an email address (no credit card). The free tier covers 60,000 characters per month; PRO at $19/mo unlocks 1,000,000.
Use the voice mapping table earlier on this page. If you still have reference audio from your old Play.ht voice, compare it against a 100-word sample generated in the suggested FreeTTS HD multilingual voice. Pick the one that fits your brand.
Replace the retired Play.ht API endpoints (api.play.ht went offline with the shutdown) with the FreeTTS public API at https://freetts.org/api. Free tier: no key required, 15 req/min. PRO+: key-authenticated, 200-1,000 req/min. Documentation and code examples for Python, Node.js, and curl are at /developers.
Run a 200-word reference script through FreeTTS and compare it against your archived Play.ht renders. Listen blind. If you cannot tell which one your audience would prefer, the migration is done — at $12/mo less than the old Play.ht entry tier.
Play.ht prices shown are the last prices the company published at play.ht/pricing before the platform shut down on December 31, 2025. The domain is now offline, so the figures are preserved here as an archival record.
FreeTTS plan details from freetts.org/pricing. Lifetime offer at freetts.org/lifetime.
FreeTTS voice count of 322 verified by GET https://freetts.org/api/voices on May 23, 2026. Marketed as "400+" including expressive style variants per voice. Play.ht 900+ figure included their PlayHT 2.0 voices plus aggregated voices from Google Cloud, AWS Polly, and third-party providers engines (archival — the service shut down December 31, 2025).
Play.ht's use of multiple synthesis backends was documented in their public technical docs and on their voice library, where engine attribution was shown per voice (both offline since the shutdown).
~1,200-1,500 characters of finished narration per minute in conversational English at default speed. Measured across 30 sample videos.
Zero affiliate links on this page. We own FreeTTS, the bias is up front, and Play.ht was recommended on merit where it won (catalog size, WordPress, streaming) while the service operated.
First three generations work without signup. PRO at $19/mo if you publish more than ~100k chars per month. Lifetime $199 if you plan to use TTS for two years or more.