Kimi K3 tops frontend code rankings but trails badly on hard math
Moonshot's Kimi K3 sits at #1 on the Code Arena Frontend leaderboard, but scores only ~39% on FrontierMath Tier 4 versus ~90% for top OpenAI and Anthropic models.
The Bench: Qwen3 14B vs Claude Fable 5 vs GPT-5.6 Sol
Three identical frontier-battery runs separated a useful local 14B model from two subscription agents, then revealed that the test needs a higher ceiling.