As of July 2026, the best text-to-speech API for most AI products is ElevenLabs: it remains the quality bar the rest of the market is measured against, with a free tier that covers building and demoing an agent before you pay. If your constraint is streaming latency rather than maximal naturalness, Cartesia is the pick; if it's unit cost at massive volume, the cloud providers win on price.
Some links below are partner links (marked). Partners never affect which tools appear, their order, or our verdicts. Same rule as our model rankings.
This roundup covers the voice layer of the BenchLM AI App Stack: the API that turns your agent's text into audio. It deliberately does not rank voice agent platforms (Retell, Vapi, OpenAI Realtime) or the LLMs behind them; that's a different decision, covered in Best LLM for Voice Agents. Pick your model there; pick its voice here.
How we compare
- Time-to-first-audio: in a live conversation, streaming latency is the product. Total generation time is almost irrelevant; when the first chunk arrives is everything.
- Naturalness under interruption: voice agents get interrupted, resume mid-sentence, and read structured data aloud. Quality that survives that is different from quality on a clean paragraph.
- Voice control: cloning, style direction, and multilingual coverage. Can you make the voice yours?
- Pricing model and free tier: character-based vs. usage tiers, and whether you can validate the build before paying.
- Lock-in: how painful switching is once your product has "a voice."
Latency gets the first criterion slot for a reason: it is the one you cannot fix later.
The field:
| Tool | Best for | Pricing model | Free tier | Standout |
|---|---|---|---|---|
| ElevenLabs | Production voice agents, voice-first products | Subscription tiers by usage | Yes — enough to demo an agent | Quality bar + cloning + streaming |
| Cartesia | Latency-critical streaming | Usage-based | Yes | Time-to-first-audio |
| OpenAI TTS / Realtime | Teams already all-in on OpenAI | Per-token/usage | Via API credits | One-vendor stack simplicity |
| Google Cloud TTS | High-volume, cost-sensitive workloads | Per character | Monthly free quota | Price at scale, language breadth |
| Amazon Polly | AWS-native products | Per character | 12-month free quota | AWS integration, price |
| PlayHT | Voice cloning breadth | Subscription tiers | Limited | Large voice library |
| Kokoro (open-weight) | Self-hosters | Your GPU bill | n/a | No per-character cost |
Prices in this category change often; check each vendor's live pricing page rather than a blog table (including ours). For the model side of the same latency budget, see LLM speed rankings.
The picks
ElevenLabs — the default
ElevenLabs (partner link) is the default recommendation for the same reason it appears in every serious voice agent tutorial: the voices survive production. Interruptions, resumed sentences, numbers and URLs read aloud, emotional register on support calls: the failure modes that make cheaper TTS feel robotic are the ones it handles. Cloning and style direction are mature, streaming is first-class, and the free tier is genuinely enough to build and test a working agent before paying anything.
Honest limits: at very high volume the per-character economics favor the cloud providers, and if your product is a phone tree reading account balances, you're paying for naturalness you don't need. That's what the scenarios below are for.
Cartesia — when latency is the product
Cartesia built its reputation on time-to-first-audio, and for live conversational agents where the human is waiting in silence, that's the metric that decides whether your product feels alive. If you've already minimized model latency with a fast LLM and you're still missing your end-to-end budget, this is the lever left.
The cloud providers — when cost is the product
Google Cloud TTS and Amazon Polly are a tier below on naturalness and a tier above on economics. For read-aloud features, notifications, accessibility, and high-volume IVR (anywhere the voice is a feature rather than the product) they're the rational pick, especially if you're already on that cloud.
OpenAI — when you want one vendor
OpenAI's TTS and Realtime APIs are the simplicity play: one API key, one bill, speech and reasoning from the same stack. You trade away voice control and the quality ceiling; you gain never thinking about the integration again.
Pick by scenario
- Production voice agent, quality-sensitive → ElevenLabs (partner link above)
- Sub-100ms streaming budget → Cartesia
- High-volume, cost-dominated (IVR, notifications) → Google Cloud TTS or Amazon Polly
- All-OpenAI stack, minimal integration → OpenAI TTS/Realtime
- Self-hosting everything → Kokoro on your own GPUs (run the self-host calculator logic on your volumes first)
Zooming out: voice is layer 5 of the AI App Stack. Upstream of it: which LLM powers the agent (latency-ranked in LLM speed) and where the whole thing runs. The stack pillar has the full map.
Reader questions
Frequently asked questions
01What is the best text-to-speech API in 2026?
As of July 2026, ElevenLabs is BenchLM's pick for production voice agents and any product where voice quality is the product — with Cartesia as the pick when sub-100ms streaming latency is the constraint, and Google/Amazon cloud TTS when cost at very high volume matters more than naturalness.
02What is the difference between a TTS API and a voice agent platform?
A TTS API converts text to audio — one layer. A voice agent platform (Retell, Vapi, OpenAI Realtime) bundles speech-to-text, an LLM, and TTS with telephony and turn-taking. If you're composing your own stack you pick a TTS API; if you want the bundle, see our voice agent guide, which compares those platforms and the LLMs behind them.
03How do I keep voice agent latency low?
Budget end-to-end, not per component: speech-to-text, LLM first-token, and TTS first-byte all add up. Choose a streaming TTS API (time-to-first-audio matters more than total generation time) and pair it with a fast LLM — BenchLM's speed rankings track first-answer latency for the model side.
04Is there a good free text-to-speech option?
ElevenLabs' free tier is enough to build and demo a working voice agent before paying. For fully free at scale, open-weight models like Kokoro exist but shift the cost to your own GPU hosting — the same build-vs-buy math as self-hosting an LLM.
Share or save