Quen
Alibaba's large language model family, including various parameter sizes that can run locally on devices.
Recently discussed in
What creators say
"Quen, that's my local model. Free unlimited intelligence because it's running on my desk. This is for like quick research. Hey, what's the top news in AI?"
More from Alex Finn →+1 more mention from Alex
"I have Quen every 20 minutes going and reading Reddit, reading Quen, searching for other people's challenges."
"If I want to use Quen 3.6 Plus, it's probably only going to cost me 33 cents per million tokens for input and about $1.95 per million tokens of output."
More from Imran Muthuvappa →"Alibaba's 35 billion parameter Quen 3.5 medium outpaces 235 billion Quen in benchmarks"
More from Peter Diamandis →Keep exploring
About this data
- ✓45,000+ mentions extracted from real podcast transcripts
- ✓Every quote links to the source episode
- ✓Sponsored mentions are always labeled
- ✓Critical takes are shown, never filtered out