@BrianRoemmele Kimi3max is legit. It tore apart a research project by Qwen loops in a good way. Sadly I can't compare to Fable because it was biology and the safeguards don't allow Farm modeling.
@BrianRoemmele It's beating Mythos? That's amazing. But expected. AI leapfrogs itself every 2 weeks. More often than that, in fact. Much more....
@BrianRoemmele Love to see it, the ridiculousness of the doom posting on display for all.
@amasad That's amazing
Many users celebrated Kimi-K3 topping the frontend code arena leaderboard and beating Claude due to its strong performance and rapid progress via distillation, while some directed frustration at Anthropic's team or marketing.
Based on 19 visible X reactions from 125 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@BrianRoemmele Kimi3max is legit. It tore apart a research project by Qwen loops in a good way. Sadly I can't compare to Fable because it was biology and the safeguards don't allow Farm modeling.
@BrianRoemmele It's beating Mythos? That's amazing. But expected. AI leapfrogs itself every 2 weeks. More often than that, in fact. Much more....
@BrianRoemmele Love to see it, the ridiculousness of the doom posting on display for all.
@amasad That's amazing
@amasad This is racist
Kimi K3 is a very good model, but people are overindexing on an Arena score again (remember Llama 4?) ELO scores as judged by Arena users are limited, and front-end is like text chat, relatively easy to train/system prompt to a state that people prefer when it is subjective.
I was personally really impressed by how they could distill fable 5 so well that it beats fable on this benchmark
The model outperforms Claude Fable 5 at one-third the cost.
@amasad This is racist
Kimi K3 is a very good model, but people are overindexing on an Arena score again (remember Llama 4?) ELO scores as judged by Arena users are limited, and front-end is like text chat, relatively easy to train/system prompt to a state that people prefer when it is subjective.
I was personally really impressed by how they could distill fable 5 so well that it beats fable on this benchmark
KImi K3 is now beating the Mythos and you know what? No drama. No fear theater. Just an AI model release. Anthropic you are ridiculous.
Apparently the distillation model can outperform the teacher model 😂
the best model in the world (for design) is open source what a time to be alive.
"They just distill" had always been such insane cope
Apparently the distillation model can outperform the teacher model 😂
Big news: Kimi-K3 by @Kimi_Moonshot is now #1 in the Frontend Code Arena with 1679 pts, surpassing Claude Fable 5. This is a 17-place jump from Kimi-k2.6 (#18 -> #1). In Frontend, Kimi-K3 ranked #1 in 6 of 7 domains: Brand & Marketing, Reference-Based Design, Data & Analytics, Consumer Product, Simulations, and Content Creation Tools, landing #2 only in Gaming behind Fable 5. The full model weights will be released by July 27. Congrats to the @Kimi_Moonshot team on this major milestone!
they distilled our models that we didn't invented yet https://x.com/birdabo/status/2039047493770752006/video/1
Big news: Kimi-K3 by @Kimi_Moonshot is now #1 in the Frontend Code Arena with 1679 pts, surpassing Claude Fable 5. This is a 17-place jump from Kimi-k2.6 (#18 -> #1). In Frontend, Kimi-K3 ranked #1 in 6 of 7 domains: Brand & Marketing, Reference-Based Design, Data & Analytics, Consumer Product, Simulations, and Content Creation Tools, landing #2 only in Gaming behind Fable 5. The full model weights will be released by July 27. Congrats to the @Kimi_Moonshot team on this major milestone!
insane. they distilled american frontier models that haven't even been invented yet. damn chinese
Big news: Kimi-K3 by @Kimi_Moonshot is now #1 in the Frontend Code Arena with 1679 pts, surpassing Claude Fable 5. This is a 17-place jump from Kimi-k2.6 (#18 -> #1). In Frontend, Kimi-K3 ranked #1 in 6 of 7 domains: Brand & Marketing, Reference-Based Design, Data & Analytics, Consumer Product, Simulations, and Content Creation Tools, landing #2 only in Gaming behind Fable 5. The full model weights will be released by July 27. Congrats to the @Kimi_Moonshot team on this major milestone!
Apparently the distillation model can outperform the teacher model 😂 https://twitter.com/arena/status/2077824029126504525
"They just distill" had always been such insane cope
China is now leading the AI arms race after spending a tenth what the US did https://twitter.com/arena/status/2077824029126504525
I was personally really impressed by how they could distill fable 5 so well that it beats fable on this benchmark
It turns out that Kimi 3 has been trained on the English language, which was an idea that China stole from anthropic (anthropic trained Claude on the English language)
Many users celebrated Kimi-K3 topping the frontend code arena leaderboard and beating Claude due to its strong performance and rapid progress via distillation, while some directed frustration at Anthropic's team or marketing.
Based on 19 visible X reactions from 125 accounts; directional sample.
Ask a question below.
Published answers will appear here.
KImi K3 is now beating the Mythos and you know what? No drama. No fear theater. Just an AI model release. Anthropic you are ridiculous.
Apparently the distillation model can outperform the teacher model 😂
the best model in the world (for design) is open source what a time to be alive.
"They just distill" had always been such insane cope
