Praney Behl

Building Sustainable AI Products Solo | 20 Years Software Engineering | Voice AI, Devtools, Workflow Autom

[Praney Behl @praneybehl]

Seriously

[Praney Behl @praneybehl]

@Sauers_ Sure

[Praney Behl @praneybehl]

@AlexEngineerAI The hidden cost in audiobook work is iteration. A 3-hour book is not just 3 hours of audio. It is pickups, pacing fixes, pronunciation, alternate reads, and mastering passes. Did you try #voisSo yet? Try free @ https://t.co/XOHMyKb28Y

[Praney Behl @praneybehl]

Your scripts never leave your machine. Vois generates voice locally on your desktop, with no uploads of your text and no per-character meter on paid plans. Built for retakes, not rationing. https://t.co/xbPRUc7CQT

New discounts now available, limited to 100 founders only.

[Praney Behl @praneybehl]

@sivori @ElevenLabs Long narration is exactly where credits get painful. It is never one clean generation: voices, retakes, pacing fixes, and chapter edits all compound fast. Try #voisSo https://t.co/WJixyFU3A3 and stop counting tokens/credits vanishing

[Praney Behl @praneybehl]

@srikrishnadas @ElevenLabs That is the worst part of metered voice tools: failed attempts still feel billable. Local generation changes the loop because retakes and experiments stop burning credits. Tried #vois https://t.co/WJixyFU3A3 yet?

[Praney Behl @praneybehl]

@theo Same!

[Praney Behl @praneybehl]

@SaidAitmbarek Too late

[Praney Behl @praneybehl]

@theo I was still on the $200 plan back then, and it never felt unlimited.

[Praney Behl @praneybehl]

Anthropic just dropped Fable 5 — the first real Mythos-class moment.

And OpenAI’s silence is becoming part of the story.

Is this the point where Anthropic pulls ahead… or is OpenAI waiting to fire back with something bigger?

Fable 5 vs OpenAI. Who’s winning?

[Praney Behl @praneybehl]

@thsottiaux OpenAI don't seem to have an answer to Fable/Mythos 5. Code RED?

[Praney Behl @praneybehl]

@uiux_harshit It's definitely a prompting-skill issue. You can start here: https://t.co/uNbwQC33EO

[Praney Behl @praneybehl]

Free to download for macOS (Apple Silicon) and Windows. Run it on your own machine and tell me what you make.

https://t.co/OXiXvSFRVe

[Praney Behl @praneybehl]

Vois 2.0 is here! 🚀 The desktop AI Voice studio that runs entirely on your machine just got its biggest upgrade: a rebuilt voice library, 600+ languages on-device, and a way to design a voice from text. No cloud. No per-character fees. 👇👇

[Praney Behl @praneybehl]

Omni and Voice Design are on the new Pro plan ($49/mo or $264/yr); everything else is on Subscriber ($29/mo or $192/yr). Omni needs a compatible GPU, so the free 7-day trial includes it. Test your hardware first.👇👇

[Praney Behl @praneybehl]

Under it all: three engines (a fast English engine, an expressive one with inline emotion tags, and a multilingual one for 23 languages), faster generation, and unlimited use offline.👇👇

[Praney Behl @praneybehl]

Voice Design lets you describe a voice in plain words (gender, age, accent, pitch, style) and Vois builds it. No recording, no digging through a library. Voice cloning is built in too, from a short sample, processed locally. 👇👇

[Praney Behl @praneybehl]

Omni is the new flagship model listing. Text to speech in 600+ languages, generated right on your machine with GPU acceleration. Your scripts never leave your device, and there is no usage meter ticking while you work. 👇👇

[Praney Behl @praneybehl]

Start with the voices. 100+ studio voices across 21 categories, from narrators and game characters to podcast hosts and app assistants. Warmer and more natural than our old set, and every paid plan gets all of them. 👇👇

[Praney Behl @praneybehl]

@addyosmani @googlecloud Nice one, is there a link to the Google's Agent factory?