Thread

1. bearjaws · 1790205736 · focus · HN ↗

   If you care about speed Cerebras gpt-oss-120b is 1400tk/s and "just as smart" in ranking.

   I've used it on a few for fun projects and its decent but the speed is crazy to watch.

   1. scosman · 1790211067 · focus · HN ↗

      Or better: Qwen 2.8 27b

      1. RussianCow · 1790212038 · focus · HN ↗

         Unfortunately, the lack of an input cache discount makes it prohibitively expensive for most use cases that aren't one-shot prompts.

         1. scosman · 1790221742 · focus · HN ↗

            well same applies to GPT OSS 120. Qwen is just the much smarter model of the 2 public options on Cerebras.

Unofficial Hacker News client; not affiliated with Y Combinator.