Model · Voice and speech

Dia

A free, downloadable speech model that turns a written two-person script into spoken dialogue, including laughs and coughs. English only. It can copy a voice from a short clip. You need a graphics card and some technical skill to run it.

Go to Dia’s website

At a glance

Open or closed: Fully open
Weights and code downloadable under a standard open licence
Code and weights released under Apache 2.0.
source · checked 12 Oct 2026 · checked on the maker's own page
Licence: Apache License 2.0 (Permissive (e.g. Apache 2.0, MIT))
README disclaimer also asks users not to make audio resembling real people without permission, or deceptive or illegal content.
source · checked 12 Oct 2026 · checked on the maker's own page
Can I sell what I make?: Yes
Follows from the Apache 2.0 licence; the README disclaimer restricts deceptive or non-consensual imitation, not commercial use.
source · checked 12 Oct 2026 · checked on the maker's own page
Cost: Free option: Free to download; you pay for your own hardware. A free online demo runs on Hugging Face.
Plans below.
How you use it: On your own computer, Through another service
Weights at huggingface.co/nari-labs/Dia-1.6B-0626; demo Space on Hugging Face. CPU support 'to be added soon'.
source · checked 12 Oct 2026 · checked on the maker's own page
Skill needed: Technical
Installed with Python and PyTorch on a CUDA GPU; the Hugging Face demo needs no setup.
source · checked 12 Oct 2026 · checked, but the page was unclear or partial
Made by: Nari Labs
The repository does not state a country.
source · checked 12 Oct 2026 · checked on the maker's own page
Status: Available
source · checked 12 Oct 2026 · checked, but the page was unclear or partial

What you can make with it

A voice or speech, read my text aloud (main use)
A voice or speech, a copy of my own voice

text to audioaudio to audio

Generates multi-speaker dialogue from a transcript; voice cloning from a 5-10 second audio prompt plus its transcript.

source · checked 12 Oct 2026 · checked on the maker's own page

Plans and prices

No paid plans listed.

source · checked 12 Oct 2026 · checked on the maker's own page

Rights and safety

Real faces and voices: Copies a real voice (audio prompt clone), no consent step
Any 5-10 s clip can be used; only a written disclaimer asks for permission, with no check in the software.
source · checked 12 Oct 2026 · checked on the maker's own page

Where you can run it

Hugging Face

Other options for “A voice or speech, read my text aloud”

Adobe Firefly, Akool, Amazon Polly, Arcads, Async (formerly Podcastle), Azure AI Speech, Bark, CAMB.AI, Cartesia, Chatterbox, Clipchamp, Deepgram, DeeVid AI, Descript, Dubverse, Eleven v3, ElevenLabs, F5-TTS, fal, Filmora, Fish Audio, FlexClip, Fliki, getimg.ai

About this entry

Facts last checked 12 Oct 2026 against the maker’s own pages (each fact links to its source). Entry status: draft. Plain-English summary written by Plainly.

Dia