HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by scosman | Hacker News Reader
Parent
Full thread
scosman
·
Or better: Qwen 2.8 27b
View on HN
RussianCow
·
Unfortunately, the lack of an input cache discount makes it prohibitively expensive for most use cases that aren't one-shot prompts.
scosman
·
well same applies to GPT OSS 120. Qwen is just the much smarter model of the 2 public options on Cerebras.
Reply on news.ycombinator.com