OpenAI.fm offers an interactive developer demo for the text to speech model in OpenAI’s API.
Zonos2
Generate speech from text with Zonos2, an open-weight mixture-of-experts model.
About Zonos2
Zonos2 is an open-weight text-to-speech model distributed as a GitHub repository. It turns text into speech and supports voice cloning from a reference audio file. You can run it through a local server with a web UI, a POST endpoint, or an offline Python API. High-performance GPU inference is documented for Linux on x86_64 with an NVIDIA GPU and a CUDA toolkit matching the driver; installation uses uv. A ggml implementation is included for CPU and cross-platform inference.
Pricing is not clearly stated.
Key features
- TTS inference server built on Mini-SGLang
- Streaming POST endpoint automatically splits long text into overlapping chunks
- Web UI can be prepopulated with voice-clone speakers from disk
- Server and offline API expose matching conditioning controls
- Text normalization can be turned off to pass text through verbatim
Best for: Local text-to-speech inference on Linux with an NVIDIA GPU, or CPU and cross-platform inference through ggml.
Good for
- MIT-licensed repository.
Zonos2 at a glance
- Website
- github.com/Zyphra/Zonos2
- Category
- AI & Machine Learning › AI Speech Synthesis
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Details verified
- Oct 22, 2026 · Official source
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
Zonos2 alternatives
Tools and sites you could use instead, ranked by how closely they match.
VoiceCraft supports zero-shot speech editing and text-to-speech generation.
TTS Online converts text into speech with more than 100 voices in 27 languages and MP3 downloads.
Free
24 alternatives
Voice.ai provides AI voice agents, a voice changer and text-to-speech tools, with APIs and SDKs for scalable speech applications.
Freemium
24 alternatives
Balabolka offers freeware text to speech software for converting written text into spoken audio.
Free
24 alternatives
Voice Dream Reader uses text-to-speech to read PDFs, web pages, and books aloud.
Free trial
24 alternativesPopular in AI Speech Synthesis
Well-known listings from the same topic.
Speechifyspeechify.comSPEECHMAspeechma.comElevenReaderelevenreader.ioVanillaVoicevanillavoice.comPaper2Audiopaper2audio.com
Zonos2 FAQ
Which languages are listed for the model?
Tier 1: English, Mandarin Chinese, Japanese. Tier 2: Korean, Russian, Italian, Portuguese, French, Spanish, Vietnamese, German, Hebrew, Dutch. Tier 3: Swedish, Hindi, Tamil, Telugu, Thai, Norwegian, Bengali, Tagalog, Arabic, Danish, Indonesian, Polish, Ukrainian, Romanian, Finnish, Hungarian, Lithuanian, Estonian, Slovak, Croatian, Latvian.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.