Jimmy Chen at Assembly AI. Voice AI Infrastructure!
San Francisco
Joined October 2026
- Tweets27
- Following74
- Followers9
- Likes96
Jimmy Chen retweeted
@AssemblyAI Universal 3.6 Pro has the lowest WER on clean audio: 1.77%. Their final transcript lands 91ms after the caller stops
Over 400 scoops of ice cream in San Francisco today! #VoiceAI turns out there was some conference nearby @AssemblyAI
Jimmy Chen retweeted
Introducing AssemblyAI-TTS-1, our first text-to-speech model.
Natural speech for voice agents, streaming in 90 ms, at $15 per million characters. On accuracy, naturalness, and cost together, it’s the best TTS model you can run at scale.
Jimmy Chen retweeted
AssemblyAI-TTS-1 is live today: our first text-to-speech model.
On accuracy + naturalness + cost, no other TTS model is better.
Frontier quality, 80ms TTFB, instant voice cloning, 4,000+ voices, unlimited concurrency, and just $15 per 1M characters.
More info 👇
From the great work by the team at @myhamletinc !
Jimmy Chen retweeted
We had so much fun at SF Tech Week that we're doing it again 👀
On October 6 we parked an ice cream truck in downtown San Francisco and handed out free scoops to the Voice AI community. Hundreds of you grabbed a scoop and chatted with us about the future of Voice AI.
So the truck is coming back.
🍦 Thursday, October 29
🕛 12:00 to 5:00 PM PT
🌉 Downtown SF (RSVP to get the address, one location this time)
Free ice cream. Bring your friends.
RSVP: luma.com/patxd4yk
Jimmy Chen retweeted
I’ll recommend something backed by data:
- Apple Dictation for Nigerian English has a 65 % word error rate.
- Blurt (open‑source, powered by AssemblyAI) has a 17 % error rate.
We hope so! Hamlet does a lot of really cool stuff where you can go and see specific profiles of your elected official. Think like seeing everything your local school board member thinks or what their record is / what they stand for.
@myhamletinc @subes01
The most important thing you can do is to get out of the way of the pictures / kiss. Side step the hell away, and fast!
- Rev. Jimmy Chen
Jimmy Chen retweeted
@covaldev's latest streaming Speech-to-Text benchmark:
Universal-3.6 Pro has the lowest WER of any model tested.
Median time to final transcript dropped about a third this week:
3.6 Pro: 153 → 106 ms
3.5 Pro: 156 → 108 ms
Thanks to the Coval team for running these benchmarks in public every week.
Jimmy Chen retweeted
Gemini 4 Argon is live on LLM Gateway.
On Voice Standard Bench it posted the top dictation score of 50+ models (99.5) and returned faster than Gemini 3.8 Flash at every transcript length.
Model string: gemini-4-argon
Full results: assemblyai.com/blog/gemini-4…
The spookiest and sugariest drink this side of FiDi. I think I almost died trying to chug this.
Where do I send my address for the iPhone Duo???
@WillowVoiceAI
@AssemblyAI
