Kimi K3-256k

For me the sweet spot is somewhere under 500k depending on how extensive I want to get. You can build up a sizable effort project in half a million tokens with Claude, with Claude having all the context from ground 0 to wherever you're off at.

Re: Kimi K3-256k

#5

Post by HawtAds » Wed, Jul 29, 2026, 7:45 PM UTC

This is just an API level change right? The model itself should be the same I think.

Re: Kimi K3-256k

#6

Post by cyanydeez » Wed, Jul 29, 2026, 7:46 PM UTC

giancarlostoro wrote: post #4

madihaa wrote: post #2

[flagged]

For me the sweet spot is somewhere under 500k depending on how extensive I want to get. You can build up a sizable effort project in half a million tokens with Claude, with Claude having all the context from ground 0 to wherever you're off at.

I'm always curious what you guys are working on; every git repo I've run a local model on and stick below <100k to increase speed seems effective enough to scope patches and changes.

Re: Kimi K3-256k

#7

Post by wxw » Wed, Jul 29, 2026, 7:52 PM UTC

> k3-256k is now available. Within 256k context, it delivers the same results. k3 (1M) consumes about twice as much quota as k3-256k.

Re: Kimi K3-256k

#8

Post by dgritsko » Wed, Jul 29, 2026, 7:52 PM UTC

This isn't quantized, right? Just a smaller context?

Re: Kimi K3-256k

#9

Post by ibuildproducts » Wed, Jul 29, 2026, 7:56 PM UTC

omg! new model!!

Re: Kimi K3-256k

#10

Post by DSingularity » Wed, Jul 29, 2026, 8:03 PM UTC

dgritsko wrote: post #8

This isn't quantized, right? Just a smaller context?

Its 256k context window. Quantization is orthogonal. We cant really tell directly so it could be quantized.