Kimi K3-256k: Mastering Model Switching and Context Limits
I explain how to navigate the new Kimi K3 and Kimi K2.7 Code models, detailing their specific context windows and speed tiers. You will learn why switching models invalidates your context cache and how to avoid unnecessary token costs by starting fresh sessions. I also clarify common errors like 401 responses and why HighSpeed mode might not feel faster when tool execution dominates the workflow.
Start a new session when using the new model: this gives better results and lower consumption.
- wren6991
This seems functionally similar to OpenAI having a step in pricing once you exceed a certain context length (also at 272k aka 2^18 aka 256k).
Having a lot of active context increases the per-token cost (flops issued and bytes read per token out) so it makes sense to pass that cost on to users. I'm actually surprised it's implemented as a hard cutoff instead of a smooth gradient.
- xyzsparetimexyz
Wow. So kimi is suddenly half the price for all users until they hit 256k of context? Thats massive.
- hawtads
This is just an API level change right? The model itself should be the same I think.
- wxw
> k3-256k is now available. Within 256k context, it delivers the same results. k3 (1M) consumes about twice as much quota as k3-256k.
- dgritsko
This isn't quantized, right? Just a smaller context?
- try-working
i've never had any issues with 256k context. see no reason to bump up to 1m if it comes at a premium.
- lukan
Since Claude is the first time for me really, really out (TIL against my wished about https://status.claude.com/), I am now interested enough to see what else works. But ... when I click pricing, I see "Join a waitlist". Wtf? Are they really that good, so were totally surprised and overwhelmed by the requests, is this a marketing stunt, or do they just don't have the hardware being in china?
- timcobb
Codex uses 256k masterfully, 1M is luxurious but still quite expensive and seems not necessary as a default.