#23
Kimi K3
kimi-k3Helpfulness
Instruction
Following
Comprehension
Empathy
Creative
Writing
Helpfulness
0.0
Instruction Following
0.0
Comprehension
0.0
Empathy
0.0
Creative Writing
0.0
Speed
Avg 58 tok/s
Overall Assistant Score
An average score combining the 5 main categories.
90.60 pts
Rank #23
78th Percentile
0
Novice
33
Capable
66
Proficient
100
Expert
Kimi K3 is Moonshot's July 2026 open frontier flagship (2.8T MoE, 1M context, always-on reasoning) for long-horizon work—coding agents, research, and messy multi-step help. Our tests put it near the top of open assistants on helpfulness, instruction following, and comprehension, with strong tool follow-through. Empathy is capable but cooler than Claude or Grok. Creative writing pays the reasoning tax: default max thinking favors agent scratchpads over sustained voice, so our tests still prefer Claude Fable and the old GPT-4.5 peak for prose. Moonshot itself notes everyday UX still trails Fable 5 and GPT-5.6 Sol. Best open pick when you want an agent that finishes hard work; not the warmest writing partner. Scored at cost-efficient reasoning (not max showcase).
Overall Score vs Price
Price axis is logarithmic (USD per 1M input tokens) · Higher score and lower price is better
Ringed point is this model
Anthropic
Baidu
Bytedance
DeepSeek
Google
Meta
Minimax
Mistral
Moonshot
OpenAI
Qwen
SpaceX AI
Tencent
Xiaomi
Zhipu AI
Compare this model
Intelligence
Overall Score · Higher is better
Speed
Output tokens per second · Higher is better
Price
USD per 1M input tokens · Lower is better
Moonshot Models
Overall Score · Same provider
Moonshot Family
Overall Score · Same model family
Closest Rivals
Overall Score · Nearest by overall rank