Getting a voice agent to read a user's emotion used to be a pipe(cat) dream. Today’s multimodal decision models, like Google Decision 1, make it easy.
But they’re large general-purpose models that read tone as one task among many. A bit overkill, and it can add 100–200ms to each turn 😨
This is why we built Pipecat Voice Reader, a tiny open-weights audio model that runs on CPU and streams results in as little as 25ms.
Here it is head-to-head with Decision 1. It holds its own on accuracy, which is crazy for a 28MB int-8 model (or 14mb for the tiny version.) Pair it with Pipecat Tone Reader, which reads the transcript, and you get a much cleaner signal on sarcasm vs. sincerity.
Demo code is available here:
https://t.co/ZnLVrGL3YG
Had a blast judging the @tryvexoai hackathon this weekend at YC. Saw some incredible voice + video hardware projects powered by @DeepgramAI and @pipecat_ai
Incredible work by the @tryvexoai team
You might've heard of @OpenAI, @OpenRouter, @opencode
We're excited to launch @opentag : the model-agnostic Claude Tag.
If you've seen Claude Tag, you know the interface. Tag an AI in Slack, hand it a task, get the result in the thread. We think that's the right interface.
We don't think the multiplayer harness should be locked in to a model provider. When a better model ships, your coworker should get better the same day.
Started early access 2 weeks ago. Live with 10 teams.
To celebrate the launch: 2 months free if you like and comment on this post. We'll DM you.
https://t.co/BSm3Smk2fI
been texting with a friend all week. normal texts. memes. weekend plans. etc
turns out i was talking to his OpenClaw agent the entire time...
he trained it to sound exactly like him, and apparently it saves him 5+ hours a week
here's how he did it:
been texting with a friend all week. normal texts. memes. weekend plans. etc
turns out i was talking to his OpenClaw agent the entire time...
he trained it to sound exactly like him, and apparently it saves him 5+ hours a week
here's how he did it:
but here's what made it scary good — he wired in live context
calendar access. location. actual life events
so when i said "how was your weekend" it knew
Sim Studio (@simstudioai) is an open-source platform to build AI agents.
Connect LLMs to your favorite tools (Slack, Pinecone, Supabase) to build agent workflows in minutes — not days.
https://t.co/TGLrNy04QR
Congrats on the launch, @emkara and @typingwala!
LET'S PACK!!!
We packed 12 pieces of apparel in 8 minutes with our mobile manipulator, Simba V1.
@DeftRobo picks e-commerce items like t-shirts from a tote cart, scan their barcodes, and pack them into an auto bagging machine.