Dia
AI
text-to-speech model
SHOW HN2025
652points
190comments
#45of 3,837 in 2025
Top 2%of 2025 by points
How HN took it
MixedNo clear lean
36 positive, 50 neutral or asking, 8 mixed, 11 critical, out of 105 comments about the project. Thread score +0.26.
Performance: 10 for, 4 againstExecution: 7 for, 6 againstIdea: 7 for, 0 againstDocs: 3 for, 1 againstDesign: 1 for, 0 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Audio quality and expressiveness sound realistic and human
From 3 comments
Sounds really good & human! Got a fair bit of unexpected artifacts though. e.g. 3 seconds hissing noise before dialogue. And music in background when I added (happy) in an…
Havoc on HNWow first time I have felt that this could be the end of voice acting/audio book narration etc. The speed with with the ways things are changing how soon before you can make any…
xbmcuser on HNI'm lost for words. This is extremely impressive!
flashblaze on HN
Impressive for a two-person team with no prior speech experience
From 2 comments
This is a pretty incredible three month creation for a couple of people who had no experience with speech models.
llm_nerd on HNHey, this is really cool! Curious how good the multi-language support is. Also - pretty wild that you trained the whole thing yourselves, especially without prior experience in…
codingmoh on HN
Objections
English only
From 3 comments
Incredible quality demo samples, well done. How's the performance for multilingual generation?
isoprophlex on HNImpressive! Is it english only at the moment?
brumar on HNIs this English-only? I'm looking for a local model for Finnish dialogue to run.
hiAndrewQuinn on HN
Outputs are unstable with artifacts such as hissing or background music
From 1 comment
Sounds really good & human! Got a fair bit of unexpected artifacts though. e.g. 3 seconds hissing noise before dialogue. And music in background when I added (happy) in an…
Havoc on HN
Requires about 10GB VRAM, too heavy for many GPUs
From 1 comment
Darn, don't have the appropriate hardware. The full version of Dia requires around 10GB of VRAM to run. If you have a 16gb of VRAM, I guess you could pair this with a 3B param…
ivape on HN
Limited to two speakers
From 1 comment
Only two voices at the moment... We will need to upgrade the dataset to make that happen, and are considering that as one of the next steps.
toebee on HN
Requests
Multi-language support
From 3 comments
Does this only support English? I would absolutely love something like this for practicing Chinese, or even just adding Chinese dialogue to a project.
999900000999 on HNIs this English-only? I'm looking for a local model for Finnish dialogue to run.
hiAndrewQuinn on HNhey, this looks (or rather, sounds) amazing! Does it work with different languages or is it English only?
bzuker on HN
Quantized version for lower VRAM
From 1 comment
We will work on a quantized version of the model, so hopefully you will be able to run it soon! We've seen Bark from Suno go from 16GB requirement -> 4GB requirement + running on…
toebee on HN
Streaming output for lower latency
From 1 comment
Impressive demo! We'd love to use this at https://useponder.ai time to first audio is something that is crucial for us to reduce the latency - wondering if dia works with output…
sarangzambare on HN
AMD GPU support
From 1 comment
Any plans for AMD GPU support? Maybe I'm missing something, but it's not working out of the box on a 7900xtx.
oehtXRwMkIs on HN
Questions
Where the training data came from
From 3 comments
I know it’s taboo to ask, but I must: where’s the dataset from? Very eager to play around with audio models myself, but I find existing datasets limiting
dangoodmanUT on HNAmazing that you developed this over the course of three months! Can you drop any insight into how you pulled together the audio data?
gfaure on HN+1 to this, amazing how you managed to deliver this, and iff you're willing to share i'd be most interested in learning what you did in terms of train data..!
isoprophlex on HN
How to keep a consistent voice across turns
From 2 comments
Are there different voices? Or only [s1] and [s2] in the examples?
xhkkffbf on HNHow do you declare which voice should be used for a particular speaker? And can it created a cloned speaker voice from a sample?
xienze on HN
Licensing terms for research and commercial use
From 1 comment
Is this Apache licensed or a custom one? The README contains this: > This project is licensed under the Apache License 2.0 - see the LICENSE file for details. > This project…
rustc on HN
Alternatives named
ElevenLabs
From 2 comments
Impressive project! We'd love to use something like this over at Delfa (https://delfa.ai). How does this hold up from the perspective of stability? I've spoken to various folks…
stuartjohnson12 on HNFantastic model. I'm going to write a Unity plugin for this. Have been using ElevenLabs for my VR game side project, but this appears to be better
LarsDu88 on HN
Sesame CSM-1B
From 2 comments
Sounds great. Hope more language support in the future. In comparison Sesame CSM-1B sounds like trained on stoned people.
pzo on HNLooking forward to try. My current go-to solution is E5-F2 (great cloning, decent delivery, ok audio quality, a lot of incoherence here and there forcing you to do multiple…
jokethrowaway on HN
Spark TTS
From 1 comment
How does this compare with Spark TTS? https://github.com/SparkAudio/Spark-TTS
verghese on HN
Bark
From 1 comment
We will work on a quantized version of the model, so hopefully you will be able to run it soon! We've seen Bark from Suno go from 16GB requirement -> 4GB requirement + running on…
toebee on HN