@Gorden_Sun
Kimi K3，2.8T（2800B）参数，100万上下文。
根据Kimi的报告，评分仅次于Fable 5和GPT-5.6 Sol，超过Opus。接下来我要试试这个模型，看看能力是不是真的这么强。
8 Replies
0 Retweets
11 Likes
5,022 Views 
One Sentence Summary
The Kimi K3 model has 2.8T parameters and a 1M-token context; according to the report, its score ranks just below Fable 5 and GPT-5.6 Sol, surpassing Opus.
Summary
According to the official Kimi report, the K3 model reaches 2.8T parameters (280 billion), supports a 1M-token context length, and in benchmarks scores just below Fable 5 and GPT-5.6 Sol, exceeding Opus. The tweet author says they will personally test it to verify its true capability.
AI Screening
78
Influence Score 10
Published Yesterday
Language
Chinese
Tags
Kimi K3
AI Model
Large Language Model
Model Evaluation
Performance Comparison
Make your daily reading actually fit you.A daily brief built from the sources you follow. Get started free