Kimi K3 is good but it’s not cheap
Surprised that no one is offering a steep discount on the price
Whoever gets there first gets a cool $10B valuation. 😎
Kimi or the entire Chinese open source ecosystem is marching towards linear / hybrid attention. K3, despite being way larger to K2, is 2.5x more efficient.
Model arch improvement and intelligent efficiency is the way moving forward to sustain scaling law while still being affordable and eco-friendly.
Kimi K3 is now in Cursor! It scores close to the frontier on CursorBench.
It's available on US-based inference thanks to our partners Fireworks, Together, and Baseten. Zero Data Retention is also supported.
Kimi K3 is now available on ChatLLM and hosted in the US!
We have also kicked off an open-source fine-tune based on this top frontier model
This is the biggest release in the enterprise AI world! Companies can control the LLM and the data.
Congrats to Kimi 🚀🚀
Kimi K3’s weights are live - 2.8 trillion parameters, only about 50 billion active per token. That efficiency is real. But the other 2.75 trillion still have to sit somewhere the moment someone self-hosts it. Sparse compute doesn’t mean sparse memory.
Kimi K3 will be truly open-weight and 3x faster on Monday
Get ready to move all your standard workloads immediately
Anything that is on Sonnet or GPT 5.5 can move — it's faster, cheaper, and better 🚀🚀