Akshat Prakash

Most TTS benchmarks are broken. They assume long, clean, studio audio. Real users give you 2 seconds, recorded on a phone, with noise. We at @useCamb_AI built MARS8 to win there.

Hollywood isn't dead. It's evolving. And we're leading that evolution. We just shipped Koyal v2.5: The best Agentic AI filmmaking platform. It goes from your script or music to full video with consistent characters, settings & storylines. Large Production houses, music labels & ad agencies use @KoyalAI for storyboarding & pre-viz. Smaller studios use it for making content they never had the time, budget or resources to create. Write a scene. Direct the camera. Build a world. @KoyalAI brings it to life. Here's what's new: - Go from script to video: Write a scene with dialogue, pacing, emotion. Koyal fully realizes it. Using the best voice models on the planet. - Direct the camera: Push in. Pull back. Drift through a scene. Real 3D camera blocking. Shape the shot the way you see it in your head. - Build a world that stays:Lock in a location and return to it. Your scenes stay consistent from the first frame to the last. Plus: dialogue, SFX, annotations for editing, and more control over every shot. Under the hood, we benchmark 40+ models weekly with real humans so Koyal always uses the best available (more on that next week!) Our agents pick the best models for each scene in run-time so you focus on the story, we handle the complexity. You don't need to prompt a film, you can direct it. With Koyal, we're looking to replace the camera, not the filmmaker behind it. Try it now at beta [dot] koyal [dot] ai DM me for credits!

@meh_agarwal I've been struggling to find a single good tool like this. Love it! great work @koyal_ai team!

@dansemperepico @dansemperepico we live AI dubbed the entire feed vimeo.com/event/5770239

Looks like the UAE government press conference is going to be in Arabic only. I'm seeing if, in the next few minutes, Claude Code can code me something that will do live streaming text translation. Working against the clock.

In an effort to make UAE public conferences more accessible to the public in this time of challenge, we at @useCamb_AI are live-streaming it LIVE-DUBBEDI in English. Hopefully people find this helpful. vimeo.com/event/5770239

@useCamb_AI is the winner of the @AegisGrahamBell Award 2026. We were recognized in “Innovation in AI – Enterprise” ahead of some of India’s most established technology giants such as @reliancejio, @TCS, @Infosys and more As an Indian founder building from abroad, this one hits differently. We’ve been fortunate to receive recognition globally. But being acknowledged back home carries a different weight. India has everything it needs to lead in AI: talent, grit, scale, technical depth. What we sometimes lack isn’t capability. It’s narrative confidence. We know how to endure. We know how to climb. That resilience is cultural. Now we just need to be unapologetic about our wins. I'm sure our illuminating chat, @madversity will prompt you to kick off this storytelling. Happy to help. To my compatriots in Dubai and across the diaspora, stay strong in these testing times 🇮🇳 🇦🇪 The work continues. Job’s not finished. 🐍

And examples on our own docs docs.camb.ai/integrations/l…

Check us out on the Livekit official docs. docs.livekit.io/agents/models/…

MARS8 is now LIVE on @livekit You can build end-to-end voice agents & real-time apps using @useCamb_AI’s most powerful TTS stack yet. MARS8 benchmarks as the world’s leading text-to-speech engine - built for real, messy, broadcast-grade scenarios. Not lab demos. Now available on LiveKit: - MARS8-Flash (ultra-low latency, conversational) - MARS8-Pro (creative + expressive control) Want Nano & Instruct next? Drop a comment. Huge shoutout to Karlson Lee and @tom_shapland for making this happen. Docs + examples in comments - start building today. Go build something impossible. Keep the mamba 🐍

@SaidAitmbarek @MicroLaunchHQ will give it a go @SaidAitmbarek

@ackprakhack dope platform guys, upvoted feel free to launch it on @microlaunchhq too we get 30k+ makers/mo on the app

Today, we’re launching Live Dubbing for ALL. For the last 24 months, this tech has powered the world’s biggest multilingual live broadcasts: NASCAR races in Spanish, MLS All-Star games in 4 languages, even Google Demo Days translated LIVE on stage. Now, any creator, broadcaster, or startup can go multilingual in real time. Same battle-tested system that handles live chaos -- multiple speakers, crowd noise, unpredictability, while keeping you authentic in every language. 80% of the world doesn’t speak English as a first language. Your content should reach all of them. LIVE. Sign up. Go live. Reach everyone. We’re live on Product Hunt today ❤️, give us some love (link in comments) As always — mamba mentality 🐍

@baseten Start deploying MARS8-Flash on Baseten today: baseten.co/talk-to-us/mar…

MARS8 is now LIVE on @baseten . Voice agents scaling on token economics punish you for success. That changes today. With MARS-Flash on Baseten, you’re no longer limited by caps, concurrency, or tokens. You get ultra-low latency voice + 99.99% uptime, running on Baseten's global infrastructure built for real traffic, not throttled abstractions. This is how we run live, global voice systems at @useCamb_AI . Baseten’s been a key partner in helping us scale without compromising consistency under pressure. ⚡ Start with 5 hours of free MARS-Flash on Baseten; then scale on your own terms. Follow the link to get started: baseten.co/talk-to-us/mar… Keep the mamba 🐍

Why does some commentary feel electric? Peter Drury, Martin Tyler, Harsha Bhogle, and other iconic voices all have one thing in common: timing, emotion, and the right energy at the right moment 🗣️ That combination was the first and hardest challenge we set out to solve at CAMB.AI for multilingual AI commentary. It’s also what we’ve delivered across live productions with NASCAR, the LFP - Ligue de Football Professionnel, Major League Soccer, Eurovision sport, and other sports organisations Building for live environments taught us something important: moments unfold fast, and you don’t get second chances. Thus, the voice has to keep up. That’s why we built a dedicated voice model in MARS-Pro, designed for moments that happen in real time. Live sports. Entertainment. Media where energies shift quickly, and delivery matters as much as clarity. It balances emotional realism with performance, which is why it’s used for broadcasts, dubbing, and audio experiences people don’t just hear, but resonate with. Voice that feels natural, expressive, and ready for production Go build with MARS-Pro today!

Akshat Prakash (@ackprakhack), CTO at @useCamb_AI, a startup that live streams in multiple languages, says there is a risk in underinvesting rather than overinvesting in AI. “It’s not debatable that AI is going to be consequential,” he adds. “We’re not debating here whether AI is going to be the fundamental backbone of any vertical. People already believe that. The question is, is it going to be in 10, 15, or 20 years?”. Read more: fastcompanyme.com/impact/if-the-…

Voice agents don’t fail because of quality They struggle when latency gets in the way That’s why we built MARS-Flash MARS-Flash is the lowest-latency model in the MARS8 family, purpose-built for real-time voice agents and assistants. It’s optimised for speed, with TTFB as low as 150ms on Blackwell GPUs. If you’re building conversational systems, every millisecond matters and MARS-Flash is designed to respond fast enough to feel human, not scripted No extra features you don’t need. No waiting on a single model stretched too far. Go build with MARS-Flash today!

@oliviscusAI This is so old

Microsoft killed the GPU mafia 🤯 They finally open-sourced their 1-bit LLM inference framework called bitnet.cpp. It lets you run 100B parameter models on your local CPU without GPUs. - 6.17x faster inference - 82.2% less energy on CPUs 100% Open Source.

@TOEwithCurt Everything that can happen is already existing as multiple infinite branches. We are simply locomoting through branches through consciousness. This is why consciousness is so key. Mastering it means that you can have happen, any reality.

What’s your theory of consciousness?