alt Hacker NewsIt is an impressive model. Agreed on most that is written on this page, with the exception of it being fast. I ran it on my own LLM benchmark suite[1] and it is faster than DeepSeek but still much slower than leading models. But it's pricing is where it really shines.
KillSwitch-Bench 1.0
Claude Opus 5 66.9
GPT-6 Astra 57.9
Claude Fable 5.1 46.7
MiMo-V2.6-Pro 38.8
Muse Spark 1.3 36.5
Replies
ricardobeat • today at 10:52 AM
Speed seems to vary a lot with demand. Last night it was reaching 80+ tok/s
conception • today at 1:02 PM
That looks like you aren’t using the ultra speed endpoint.
➕ show 1 reply
Why does Luna have a score of 0? I will say that in my limited experience, I use Luna and DSv4 Flash (haven't tried 4.1 yet) and Luna is wayyyy faster. They both output at the same speed but DS has an endless thought process