Launching today
Earlybird
29M-param speech-to-text that outputs words before they end
14 followers
29M-param speech-to-text that outputs words before they end
14 followers
Earlybird is a 29M-parameter, 17 MB on-device speech-to-text model that predicts each word correctly about 90 ms before the speaker finishes it, with a confidence on every word. Built for voice agents that need to act sooner. Open weights and code.
Free
Launch Team
Render Sandboxes: a computer for every agentGive every agent its own machine. Starts in under a second
Promoted
Hey Product Hunt! I'm Sahil.
I've been building voice agents and kept running into the same delay: speech-to-text waits for each word to end before it shows it, so the agent can't start working until then. Earlybird outputs words before the speaker finishes speaking them. On audiobook speech the median word is predicted correctly 90 ms before it's fully spoken, where other streaming models get it right 90 to 400 ms after. The model is 17 MB and runs on a CPU.
Getting the word before it's finished is a neat trick. Does it work for languages other than English yet?