← All posts
· 8 min read

TTS API Pricing in 2026: I Went Through Every Provider So You Don't Have To

Current, verified pricing for every major TTS API in 2026 — ElevenLabs, OpenAI, Google Cloud, Amazon Polly, Azure, and LeanVox. Real numbers, real math, honest opinions.

ttsapipricingcomparisondeveloper-tools

Every time I try to compare TTS API pricing I end up on six different tabs, converting credits to characters to dollars and trying to remember whether that free tier applies before or after the other free tier. It's annoying. So here's the version I wish existed — current pricing, pulled directly from each provider's page this month, with actual math.

Fair warning: I built LeanVox, so I have skin in the game here. I've tried to be accurate and fair about the other providers anyway, because bad comparisons don't help anyone.

Prices verified February 2026. Check source pages before committing — these things change.


The quick version

If you just want the table and not the commentary:

Provider Price/1K chars Voice Cloning Languages Subscription?
LeanVox Standard $0.005 10 No
LeanVox Pro $0.01 ✓ 10s sample 23+ No
ElevenLabs (Starter, $5/mo)$0.16732Yes
ElevenLabs (Creator, $22/mo)$0.2232Yes
ElevenLabs (Scale, $330/mo)$0.16532Yes
ElevenLabs (Business, $1,320/mo)$0.1232Yes
OpenAI tts-1$0.0156No
OpenAI tts-1-hd$0.0306No
Google WaveNet / Standard$0.00440+No
Google Neural2$0.01640+No
Google Chirp 3 HD$0.030✓ Custom35+No
Amazon Polly Standard$0.00430+No
Amazon Polly Neural$0.01630+No
Amazon Polly Generative$0.030LimitedNo
Azure Neural TTS$0.016Custom Neural140+No

What this actually costs at real usage levels

Per-character rates are hard to reason about in the abstract. Here's what you'd actually pay at three different usage levels. I used each provider's best neural/natural tier for quality comparison — not the cheapest legacy voice.

Provider 500K chars/mo
(~55 min audio)
5M chars/mo
(~9 hrs audio)
50M chars/mo
(~91 hrs audio)
LeanVox Standard$2.50$25$250
LeanVox Pro$5$50$500
ElevenLabs Starter~$83 + $5/mo sub~$835 + $5/mo sub~$8,350 + sub
ElevenLabs Scale~$82 + $330/mo sub~$825 + $330/mo sub~$8,250 + sub
OpenAI tts-1$7.50$75$750
Google Neural2~$6 (after 1M free)~$64$784
Google Chirp 3 HD~$12 (after 1M free)~$120$1,470
Amazon Polly Neural$8$80$800
Azure Neural~$7 (after 500K free)$72$784

The ElevenLabs numbers always look shocking in tables like this. It's not that they're gouging — they're genuinely investing in quality and a subscription model is how they fund that. It's just that the math only works once you're generating revenue that justifies it.


My honest take on each provider

ElevenLabs

The quality is real. Voices sound like people, not robots. The cloning works better than any other provider I've tried. If you're building something where voice quality is literally the product — premium audiobooks, voice-over tools, branded voice assistants — and you have the revenue to support it, this is the right call.

But the subscription requirement and the per-character rate make it a terrible fit for experimentation, pre-revenue projects, or anything with variable usage. You're locked into a monthly fee whether you use it or not.

OpenAI TTS

Clean, simple, fine. Six voices, decent quality, no cloning, pay-as-you-go. If you're already deep in the OpenAI ecosystem and just need TTS without adding another vendor, this makes sense. Not a reason to switch providers just for TTS.

Google Cloud TTS

The most interesting thing about Google's TTS right now is the tier sprawl. You've got legacy Standard/WaveNet at $0.004/1K (old quality, cheap), Neural2 at $0.016/1K (solid), Chirp 3 HD at $0.030/1K (their current flagship), and now Gemini-TTS which is token-priced and hard to compare directly.

The legacy voices are noticeably dated. Chirp 3 HD is competitive quality but gets expensive. The free tier for WaveNet (4M chars/month) is the most generous free tier in the market, which makes Google the obvious choice for development and low-volume projects that can tolerate older voice quality.

The 40+ language support is genuinely hard to beat if you need it.

Amazon Polly

Similar story to Google — Standard voices are cheap and sound like 2018, Neural voices are better, Generative voices are actually decent. Free tier (5M chars/month for 12 months) is generous for getting started. If you're AWS-native and don't need voice cloning, Polly works.

I wouldn't choose Polly over the field for voice quality alone, but as part of an AWS stack it's the path of least resistance.

Azure Neural TTS

The play here is really the 140+ language count. If you need Swahili or Marathi or Welsh, Azure is probably your only realistic option. The quality is fine, the pricing is mid-tier, the integration is deep in Microsoft's ecosystem. Enterprise teams that are already on Azure will find it easy. Everyone else will find it bureaucratic to set up.

LeanVox

I built this, so take what I say with appropriate skepticism. The pitch: competitive voice quality, no subscription, credits don't expire, voice cloning at $0.01/1K. You get $0.50 to try it with no credit card — enough for 50,000 characters of Pro or 100,000 of Standard. Try it, compare, decide.

The honest limitations: English-first (23+ languages but the English voices are strongest right now), public beta, and the voice catalog is smaller than ElevenLabs. We're working on it.


How to actually decide

If you're pre-revenue and experimenting: Start with the free tiers — Google WaveNet or Polly Standard for volume, LeanVox for quality at minimal cost. Don't commit to ElevenLabs until you have users paying for the thing you're building.

If you need voice cloning specifically: ElevenLabs has the best cloning, LeanVox has the cheapest per-character rate for cloning. The question is how important the quality delta is for your use case.

If you need 40+ languages: Azure or Google. No one else is close.

If you're scaling up and ElevenLabs pricing is getting painful: This is exactly who LeanVox is for. Free credits, no card — worth trying before you sign an ElevenLabs annual contract.


Try LeanVox free

$0.50 in free credits. No credit card required. Start generating speech in 30 seconds.

Get started free →