The best text-to-speech models in 2026
The best text-to-speech models right now are Eleven v4 and v4 Turbo, Gemini 3.8 Flash TTS and Cartesia Sonic 3.6, out of 24 we have reviewed. Each is ranked by our rating, with its price, our verdict and who it suits.
By ReviewRamble editors · Updated Nov 15, 2026 · How we review
How we ranked: by our rating out of 5, from reviews built on each product's own pricing, documentation and policies plus independent benchmarks where they exist, each checked on the date shown in the review. We take no affiliate commission.
- 1
Nothing on the public leaderboards sounds better, but Google's Gemini 3.8 Flash TTS gets within 60 Elo points for about a fifth of v4's price.
Best for: Audiobook, game and dubbing producers who need the most expressive voices, and voice-agent teams who want top-ranked quality at Turbo's lower rate.
Price: API: v4 $0.08 and v4 Turbo $0.04 per 1,000 characters. Subscriptions from Starter $6/month to Business $990/month include character allowances.
Read our Eleven v4 and v4 Turbo review4.7/5 - 2
Fifth in blind votes at under half Eleven v4 Turbo's price, Gemini 3.8 Flash TTS is a bargain with an expiry date of 1 January.
Best for: Developers producing long-form narration or multilingual dubbing who want near-top quality and can absorb January's price rise.
Price: Flash TTS: $0.50 per 1M text tokens in, $9 per 1M audio tokens out (about $0.81 per hour of audio) until 31 Dec 2026, doubling on 1 Jan 2027. Flash-Lite: $6 per 1M audio tokens out. Batch half price; free tier in AI Studio.
Read our Gemini 3.8 Flash TTS review4.5/5 - 3
Sonic 3.6 closed most of the gap to ElevenLabs in one release and costs less per character than Eleven v4, but it trains on customer content unless you opt out.
Best for: Teams building production voice agents that want stable, versioned voices at a moderate price.
Price: Free plan 20K credits/month (no commercial licence); Pro $5 (100K), Startup $49 (1.25M), Scale $299 (8M); overage $38 to $65 per 1M credits. One credit is one character, about 750 to 800 per minute of audio.
Read our Cartesia Sonic 3.6 review4.4/5 - 4
ZONOS3-Mini puts multilingual voice cloning in a model small enough for a browser tab, under a licence that allows commercial use, which few open speech models offer together.
Best for: Developers who need multilingual speech or voice cloning on their own hardware, in apps, games or the browser.
Price: Free open weights under Apache 2.0, including commercial use; a hosted version is on Zyphra Cloud, with prices on Zyphra's site.
Read our ZONOS3-Mini review4.3/5 - 5
Drama 3 gives open weights a voice listeners rank alongside the paid leaders, under a licence that stops being free the moment a business or a voice API gets serious.
Best for: Small studios, game developers and app makers under the revenue cap who want near-ElevenLabs narration and multi-character dialogue on their own GPUs.
Price: Weights free under the Fish Audio Community License for research and for businesses under $500,000 annual revenue. API: $50 per 1M UTF-8 bytes, $10 until 12 November 2026.
Read our Fish Audio Drama 3 review4.2/5 - 6
VoxCPM2 is the open speech model to start with when you need a commercial licence, many languages and voice design in one download.
Best for: Product teams that need commercially usable open speech with voice design and cloning across European and East Asian languages on one GPU.
Price: Free under Apache 2.0, commercial use included. OpenBMB sells no hosted API; the cost is your own GPU time.
Read our VoxCPM2 review4.2/5 - 7
Phonon 31M is the tiny English voice to ship in an app today: smaller than Kokoro, preferred by listeners, and Apache-licensed.
Best for: App and device makers who need a natural English voice that runs offline on phones, laptops or in a browser, under a commercial licence.
Price: Phonon 31M English is free under Apache 2.0; the 40M models are CC BY-NC 4.0. Gradium's hosted API has a $0 tier with 45,000 characters a month, S at $39 a month and L at $1,480 a month for 45 million characters, on annual billing.
Read our Gradium Phonon review4.1/5 - 8
MAI-Voice-2.1 is a credible, low-priced voice for Azure teams, but it covers a quarter of the languages ElevenLabs does and has no independent quality ranking yet.
Best for: Teams already on Azure Speech who want expressive, multilingual voices at a lower price and can accept preview status.
Price: MAI-Voice-2.1 $22 per 1M characters; MAI-Voice-2.1-Flash $15 per 1M characters, through Azure Speech in Foundry; OpenRouter and Vercel list the same $22 rate.
Read our MAI-Voice-2.1 review4.0/5 - 9
For Indian-language voice agents, Maya 2 Native is the open model to start with: close to the paid leaders in Hindi and cheap to run at scale, with automated scores standing in for proof everywhere else.
Best for: Developers building voice agents, IVR or narration in Indian languages who want to self-host under a permissive licence.
Price: Free under Apache 2.0 (Mimi codec CC BY 4.0); self-hosted on your own GPUs. No first-party hosted price published.
Read our Maya 2 Native review4.0/5 - 10
OmniVoice is the open TTS model to reach for when the language is not on anyone else's list, provided your lawyer reads the tokenizer licence.
Best for: Developers who need open, self-hosted speech in languages the commercial and open leaders do not cover.
Price: Free weights; code under Apache 2.0, audio tokenizer under the Boson Higgs Audio 2 Community License (expanded licence needed above 100,000 annual active users).
Read our OmniVoice review4.0/5 - 11
Fun-CosyVoice3 is the small, fast, Apache-licensed cloner to pick for Chinese and its dialects, and a sound one for streaming agents in eight other languages.
Best for: Teams building Chinese-language voice agents or dialect content who want a small, commercially usable model they can run themselves.
Price: Free under Apache 2.0. Alibaba Cloud's hosted cosyvoice-v3-flash costs $0.13 and cosyvoice-v3-plus $0.26 per 10,000 characters internationally.
Read our Fun-CosyVoice3 review3.9/5 - 12
Nothing this small sounds this good, but it has had no updates since August 2025, so treat it as finished software rather than a project that will improve.
Best for: Developers who need cheap, permissively licensed speech for narration, accessibility or prototypes.
Price: Free under Apache 2.0. Hosted at $0.65 per million characters on Replicate (via Artificial Analysis), the cheapest on its leaderboard, and $0.99 on DeepInfra.
Read our Kokoro-82M review3.9/5 - 13
A capable open voice-cloning model with unusually broad language coverage, let down by the absence of any published quality or latency numbers and a Linux-plus-NVIDIA-only footprint.
Best for: Developers who need a self-hosted, permissively licensed voice-cloning TTS across many languages and can run it on an NVIDIA Linux box.
Price: Free weights under Apache 2.0. A hosted playground and API run at cloud.zyphra.com; API pricing is not restated on the model card.
Read our Zonos v2 (ZONOS2) review3.9/5 - 14
Audio8's 0.6B model clones voices in 23 languages at error rates near 4B-parameter rivals, a strong deal on size that still trails them on the hardest multilingual test.
Best for: Developers who need permissively licensed multilingual voice cloning on a small GPU budget and can test quality in their own languages.
Price: Free. Realistic Air and Preview 0.6B weights are Apache 2.0; Preview 0.1B is free for non-commercial use and for companies under US$2 million yearly revenue. No hosted API.
Read our Audio8 TTS review3.8/5 - 15
Breeze TTS 2 sounds like a frontier voice model, but its licence makes it a demo you can run at home, not an ElevenLabs replacement you can ship.
Best for: Researchers, hobbyists and teams evaluating open voice quality who will either stay non-commercial or buy BreezeBlue's hosted plan for production.
Price: Weights free for research and non-commercial use only. Hosted plans: Free (400 daily credits, 10 characters each), Starter $9/month, Creator $19/month, Pro $99/month, Scale $299/month, Business $999/month; about $30 per million characters, or roughly $25 billed yearly. Enterprise by quote.
Read our Breeze TTS 2 review3.8/5 - 16
Higgs TTS 3 posts the lowest error rates in Boson's own multilingual tests, but listeners on Artificial Analysis rank it mid-table and its licence keeps it out of products.
Best for: Researchers and independent creators who want accurate, expressive speech in many languages, and companies trialling Boson's API before buying.
Price: Weights free for research, non-commercial use and credited creator channels; products need a commercial licence. Boson API: $15 per million characters, $10 trial credit.
Read our Higgs TTS 3 review3.8/5 - 17
IndexTTS-2.5 is the open cloner for dubbing work where an actor's emotion has to change but their voice must not.
Best for: Dubbing and game-audio teams working in Chinese, Japanese or English who need to direct a cloned voice's emotion line by line.
Price: Free under the bilibili Model Use License, including commercial use, unless you have over 100 million monthly users or RMB 1 billion annual revenue. No hosted API.
Read our IndexTTS-2.5 review3.8/5 - 18
Indic-Speak is the safe pick for Indian-language narration at scale: consented voices, 23 languages and a cheap API, but no cloning and no independent ranking yet.
Best for: Education, government and media teams that need natural narration across many Indian languages from voices that are legally clean.
Price: Weights free under the Indic Open Model License, with credit; hosting it as an API needs Bodhan's sign-off. Hosted API: ₹6 per 10,000 characters, ₹10 free credit.
Read our Indic-Speak review3.8/5 - 19
Chatterbox is the open voice cloner a compliance team can sign off, thanks to its watermark, though not the one a listener would choose.
Best for: Teams that need commercially usable open voice cloning with provenance marking and can accept mid-table voice quality.
Price: Free under the MIT licence. Artificial Analysis lists hosted Chatterbox at $25 and Resemble's Chatterbox HD at $40 per million characters.
Read our Chatterbox review3.7/5 - 20
NeuTTS gets voice cloning onto a phone or a Raspberry Pi with a watermark attached, at the price of sounding like a small model.
Best for: Hardware and app makers who need voice cloning that runs offline on phones or embedded boards, with provenance marking built in.
Price: Air is free under Apache 2.0. Nano and 2E are free for research and for companies under $5 million annual revenue; above that a paid licence is needed.
Read our NeuTTS (Air, Nano, 2E) review3.6/5 - 21
Woof is the most accurate text-to-speech you can fit in 40 MB on Underdog's own test, and the licence keeps it out of almost any business use without a call to sales.
Best for: Hobbyists, researchers and non-profits who want a clear English voice that runs offline on any laptop CPU.
Price: Free for non-commercial use with credit under the Underdog Non-Commercial License 1.0. Commercial licence required for any business use; price not published. Free inside Underdog's own apps.
Read our Underdog Woof review3.6/5 - 22
Qwen3-TTS hands developers fine-grained voice controls and a three-second cloning tool with no guardrails, at a quality listeners rank near the bottom.
Best for: Developers building multilingual voice features who want fine control under a permissive licence and will add their own safeguards.
Price: Free under Apache 2.0. Alibaba Model Studio (Singapore): qwen3-tts-flash $0.10 and qwen3-tts-flash-realtime $0.13 per 10,000 input characters, with a 90-day free quota.
Read our Qwen3-TTS review3.5/5 - 23
KittenTTS 2 squeezes a cloning speech model into under a gigabyte, but nobody outside KittenML has yet shown how it sounds, and the licence shuts out most funded companies.
Best for: Hobbyists, researchers and unfunded developers who want offline voice cloning on ordinary hardware and can live with an unproven quality ceiling.
Price: Weights free under the Stellon Labs Community License for research and for commercial users under $1 million in revenue and in funding, with registration. Hosted API pricing not published.
Read our KittenTTS 2 review3.4/5 - 24
CSM-1B is a capable research base for conversational speech, but in late 2026 it is an older, English-only model that needs work before it sounds like Sesame's demo.
Best for: Research and product teams that want to fine-tune their own English conversational voice on an open base with a permissive licence.
Price: Free under Apache 2.0 after accepting the gate. Hosted on DeepInfra through OpenRouter at $7 per million characters.
Read our Sesame CSM-1B review3.3/5
Sources
The main official source behind each ranked review; every review lists its full set.
- Eleven v4 and v4 Turbo: ElevenLabs: Introducing Eleven v4
- Gemini 3.8 Flash TTS: Google: Gemini 3.8 Flash TTS and Flash-Lite TTS
- Cartesia Sonic 3.6: Cartesia docs: Sonic 3.6
- ZONOS3-Mini: Zyphra/ZONOS3-Mini (Hugging Face)
- Fish Audio Drama 3: Hugging Face: fishaudio/drama-3 model card
- VoxCPM2: VoxCPM2 model card
- Gradium Phonon: Phonon 31M English model card
- MAI-Voice-2.1: Microsoft Learn: MAI-Voice-2.1 and MAI-Voice-2.1-Flash
- Maya 2 Native: Hugging Face: maya-research/maya2-native model card
- OmniVoice: OmniVoice model card (Hugging Face)
- Fun-CosyVoice3: Fun-CosyVoice3-0.5B-2512 model card
- Kokoro-82M: Kokoro-82M model card
- Zonos v2 (ZONOS2): Hugging Face: Zyphra/ZONOS2 model card
- Audio8 TTS: Edge0-AI/Audio8_TTS on GitHub
- Breeze TTS 2: Breeze TTS 2 model card
- Higgs TTS 3: Higgs TTS 3 model card
- IndexTTS-2.5: IndexTTS-2.5 model card
- Indic-Speak: Indic-Speak model card
- Chatterbox: Chatterbox model card
- NeuTTS (Air, Nano, 2E): neuphonic/neutts on GitHub
- Underdog Woof: Hugging Face: ConwayResearch/Underdog-Woof model card
- Qwen3-TTS: Qwen3-TTS model card
- KittenTTS 2: Hugging Face: KittenML/kitten-tts-2 model card
- Sesame CSM-1B: CSM-1B model card
More guides
- Best large language models (LLMs)
- Best AI reasoning models
- Best AI coding models
- Best small and on-device AI models
- Best vision and multimodal AI models
- Best OCR and document AI models
- Best AI image models
- Best AI video models
- Best AI decision models
- Best speech-to-text models
- Best AI music and audio models
- Best embedding models
- Best open-source LLMs
- Best American open-source LLMs
- Best Chinese open-source LLMs
- Best European open-source LLMs