Hacker News: narrationbox

New comment by narrationbox in "Launch HN: Speko (YC S26) – OpenRouter for Voice AI"

narrationbox — Mon, 17 Aug 2026 16:24:13 +0000

> Typical production voice agent is an ensemble of three models: STT, an LLM, and TTS.

To use a claudism, I would like to push back on this. The industry is very much moving towards one-model-does-all end to end trained similar to LLMs and VLMs. Mostly for latency reasons and partially because the results for the end to end trained models are just so much better than those using three pieces architectures.

I think most of the value prop is in automatic evals, not routing specifically. A better pitch for you would be "the LM Arena of voice models" rather than comparing yourself to openrouter because the value add is rather questionable. For TTS specifically, the current SOTA for production systems are all using prompt based voice gen i.e. instead of having 10 different Tacotron models trained on 10 different models, these days it's all a single large model and the "style" is a prompt in the system prompt. The input is usually something like

  
  Speak in a deep smooth voice similar to a documentary narrator
  
  
  Speko is the ultimate evaluation platform for voice agents. We do automatic  evals.

It's the same for voice cloning too, you just pass the reference speech as an input file for all generations. A lot of systems don't have any separate style vector extraction step or model-specific fine-tuning anymore.

So something like OpenRouter for voices offer questionable value given that stakeholders usually make this sort of decisions once at the start of the project. On the other hand if you can offer automatic evals and figure out which prompts give the most similar results across different voice providers, that would offer a lot more value. It would be nice to be able to switch from e.g. Grok voice agents to ChatGPT voice agents knowing that the output style won't change too much. There are many companies now with evals as a core business model: LM Arena, Artificial Analysis, Prompt foo (before they got acquired and pivoted to security only) so many take a look at them.

Source: we have been building TTS systems for over a decade too https://narrationbox.com

New comment by narrationbox in "VibeVoice: Open-source frontier voice AI"

narrationbox — Tue, 28 Apr 2026 19:25:01 +0000

Yes, the SOTA is currently much more advanced.

New comment by narrationbox in "Eleven v3"

narrationbox — Fri, 06 Jun 2025 10:54:02 +0000

Give us a try, I think we are what you are looking for

https://narrationbox.com

New comment by narrationbox in "Google Illuminate: Books and papers turned into audio"

narrationbox — Tue, 10 Sep 2024 18:39:21 +0000

A lot of our customers use us [0] for that, it works pretty well if executed properly. The voiceovers work best as inserts into an existing podcast. If you see the articles of major news orgs like NYT, they often have a (usually) machine narrated voiceover.

[0] https://narrationbox.com

New comment by narrationbox in "Another Text to Speech API"

narrationbox — Tue, 26 Sep 2023 00:39:25 +0000

Yeah, neural codecs are pretty amazing. The most incredible part is that they can do compression well across the temporal domain, something which has been non-trivial.

New comment by narrationbox in "Project Gutenberg releases 5k free audiobooks"

narrationbox — Mon, 25 Sep 2023 01:29:04 +0000

Their sentence segmentation heuristics were not configured correctly. It's not an inherent limitation of the technology itself.

The newer transformer based generators are a bit better in this regard (since they can maintain a longer context window, not just in short tiny snippets).

New comment by narrationbox in "PlayHT2.0: State-of-the-Art Generative Voice AI Model for Conversational Speech"

narrationbox — Fri, 11 Aug 2023 20:16:12 +0000

Mel + multispeaker vocoder is very much a classic (tacotron era) TTS approach

New comment by narrationbox in "DeepFilterNet: Noise supression using deep filtering"

narrationbox — Wed, 07 Jun 2023 03:37:16 +0000

Since it does the signal processing in the Fourier domain, does this suffer from audio artefacts e.g. hissing in the output? Torch's inverse STFT uses Griffin-Lim which is probabilistic and if you don't train it sufficiently, you may sometimes get noise in the output.

https://pytorch.org/docs/stable/generated/torch.istft.html#t...

An alternative would be to use a vocoder network (or just target a neural speech codec like SoundStream).

New comment by narrationbox in "US Marines defeat DARPA robot by hiding under a cardboard box"

narrationbox — Wed, 25 Jan 2023 17:05:34 +0000

> Recognizes "human" and recognizes "desk". I sit on desk. Does AI mark it as a desk or as a chair?

Not an issue if the image segmentation is advanced enough. You can train the model to understand "human sitting". It may not generalize to other animals sitting but human action recognition is perfectly possible right now.

New comment by narrationbox in "NaturalSpeech: End-to-end text to speech synthesis with human-level quality"

narrationbox — Wed, 18 May 2022 05:04:42 +0000

Your average mobile processor doesn't have anywhere near enough processing power to run a state of the art text to speech network in real-time. Most text to speech on mobile hardware are stream from the cloud.

New comment by narrationbox in "Ask HN: Non-tech professionals on HN?"

narrationbox — Sat, 09 Oct 2021 15:55:03 +0000

For the high end stuff no, but many of the lower tier jobs are under threat.

New comment by narrationbox in "Show HN: Automated Binance Trading Bot – Buy Low/Sell High"

narrationbox — Fri, 23 Apr 2021 13:24:02 +0000

We used to be in this field too (https://kloudtrader.com/narwhal). It is a very crowded market and monetisation is tricky.

New comment by narrationbox in "Quickdraw with Google AI"

narrationbox — Mon, 29 Mar 2021 22:19:26 +0000

You can throw this together pretty quickly using one of the AutoML APIs.

New comment by narrationbox in "YouTube can now warn creators about copyright issues before videos are posted"

narrationbox — Thu, 18 Mar 2021 21:42:12 +0000

So if I am launching a Spotify/Audible-style music/audiobook streaming platform, how will the pricing work out? Do we pay you instead of the original author for any user uploaded content? If the original author chooses to upload their content onto our platform, do we scan it with your API and explicitly whitelist it and pay them directly?

Your company looks very cool btw. What's your email?

New comment by narrationbox in "Launch HN: Pry (YC W21) – Finance for Founders"

narrationbox — Thu, 18 Mar 2021 21:39:29 +0000

What about TransferWise ewallets?

New comment by narrationbox in "YouTube can now warn creators about copyright issues before videos are posted"

narrationbox — Thu, 18 Mar 2021 21:38:20 +0000

Do you have any public pricing or startup plans?

New comment by narrationbox in "Launch HN: Pry (YC W21) – Finance for Founders"

narrationbox — Tue, 16 Mar 2021 17:58:09 +0000

Does the accounting system support Canada and other commonwealth countries? Or is this US only?

New comment by narrationbox in "Launch HN: Enombic (YC S20) – Create your own stock indexes"

narrationbox — Thu, 25 Feb 2021 19:30:26 +0000

Some ETFs also allow easy investment in stocks of foreign countries without requiring additional brokerages on behalf of the end user. I presume this does not offer that functionality?

New comment by narrationbox in "Show HN: Turn scripts into fine-tuned voices via Wiki markups"

narrationbox — Tue, 23 Feb 2021 19:58:57 +0000

I will look into it, Wiki2SSML looks very handy.

New comment by narrationbox in "Show HN: Turn scripts into fine-tuned voices via Wiki markups"

narrationbox — Mon, 22 Feb 2021 04:02:20 +0000

It looks great, have you considered adding a visual editor?

We have one for our systems: https://narrationbox.com