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.
My current Claude Code session has been going on for like 35 hours and has used up around 400 million tokens, thankfully almost all of those being cached (95-98%) - pretty typical for long form agentic work.
First you spend like 2-3 hours working on a plan, once you have that you just tell the model to go and implement it, do adversarial sub-agent review loops before each commit and also make sure that all tooling and tests pass (including coverage requirements). You do need to poke it in a slightly different direction every few hours, though. Not even any novel work, just some refactoring and SSE notification hardening, bug fixes, alongside environment tuning and getting rid of some bottlenecks (also migrated from Oracle to PostgreSQL but that's mostly done).
That said, Kimi somehow manages to use less context in the main thread than Anthropic's models (even when you use sub-agents and also dynamic workflows in Claude Code), might have something to do with either how the model is tuned or their Kimi Code harness - because even in most of the longer form sessions it doesn't seem to fill up quite as quickly (note: because the kimi vis tool doesn't have a full summary view across all agents, these are the main long running agent stats across some sessions, not sub-agents):
I could see 256k context being sufficient for all sorts of work, even if intermediate progress/plan tracking files and docs might have to be used along the way, in addition to whatever plan support the harness has (for example, if you document something that will be relevant for load testing you might need that in 10 turns but not during the ones before then).
They are just talking to the model in CC, while staying in a single thread. Doubt they have any actual coding knowledge to compartmentalize different problems in the codebase.
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?
Kimi had become that popular. I was a subscriber of Kimi back when latest version was Kimi K2. Later I unsubscribed because I jumped over to GLM subscription (they had amazing deal). Now when I wanted to try out Kimi K3 to find out what the fuzz was all about, I couldn’t subscribe to them.
I remember reading a post from Moonshot team about this, they are doing this because they are almost at peak capacity and want to reserve it to keep the quality for their current customers.
We are actually witnessing an open-weight model catching up at catching mainstream users attention. And instead of behaving like Anthropic, they actually care about their users experience.
It's an extremely popular model hosted by a company affected by hardware export bans. I doubt they'd voluntarily prevent people from subscribing if they didn't absolutely have to to maintain service quality.
As a user, I much prefer that to service disruptions or severely degraded or secretly quantized performance. However if I didn't have an account, I'd be pretty pissed off about not being able to give them money and become a user.
This was posted 38 minutes ago, and as of 20 minutes ago, several Anthropic services are now designated as having a "major outage".
Doubt these are related, but it made me laugh a little.
This is just an API level change right? The model itself should be the same I think.
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.
> k3-256k is now available. Within 256k context, it delivers the same results. k3 (1M) consumes about twice as much quota as k3-256k.
That's actually nice! I usually try to stay below 200k context anyway.
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.
My current Claude Code session has been going on for like 35 hours and has used up around 400 million tokens, thankfully almost all of those being cached (95-98%) - pretty typical for long form agentic work.
First you spend like 2-3 hours working on a plan, once you have that you just tell the model to go and implement it, do adversarial sub-agent review loops before each commit and also make sure that all tooling and tests pass (including coverage requirements). You do need to poke it in a slightly different direction every few hours, though. Not even any novel work, just some refactoring and SSE notification hardening, bug fixes, alongside environment tuning and getting rid of some bottlenecks (also migrated from Oracle to PostgreSQL but that's mostly done).
That said, Kimi somehow manages to use less context in the main thread than Anthropic's models (even when you use sub-agents and also dynamic workflows in Claude Code), might have something to do with either how the model is tuned or their Kimi Code harness - because even in most of the longer form sessions it doesn't seem to fill up quite as quickly (note: because the kimi vis tool doesn't have a full summary view across all agents, these are the main long running agent stats across some sessions, not sub-agents):
I could see 256k context being sufficient for all sorts of work, even if intermediate progress/plan tracking files and docs might have to be used along the way, in addition to whatever plan support the harness has (for example, if you document something that will be relevant for load testing you might need that in 10 turns but not during the ones before then).They are just talking to the model in CC, while staying in a single thread. Doubt they have any actual coding knowledge to compartmentalize different problems in the codebase.
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?
No, the waiting list is true.
Kimi had become that popular. I was a subscriber of Kimi back when latest version was Kimi K2. Later I unsubscribed because I jumped over to GLM subscription (they had amazing deal). Now when I wanted to try out Kimi K3 to find out what the fuzz was all about, I couldn’t subscribe to them.
I remember reading a post from Moonshot team about this, they are doing this because they are almost at peak capacity and want to reserve it to keep the quality for their current customers.
We are actually witnessing an open-weight model catching up at catching mainstream users attention. And instead of behaving like Anthropic, they actually care about their users experience.
It's an extremely popular model hosted by a company affected by hardware export bans. I doubt they'd voluntarily prevent people from subscribing if they didn't absolutely have to to maintain service quality.
> Are they really that good,
They did exceed the expecations of pretty much everyone! I've also blogged about using the model, it's a bit on the slow side but pretty good!
> so were totally surprised and overwhelmed by the requests ... or do they just don't have the hardware being in china?
Yes, this is mostly the case: https://x.com/Kimi_Moonshot/status/2078855608565207130
As a user, I much prefer that to service disruptions or severely degraded or secretly quantized performance. However if I didn't have an account, I'd be pretty pissed off about not being able to give them money and become a user.
Since the model is open-weights, you can get from other providers, for example see https://openrouter.ai/moonshotai/kimi-k3#providers
omg! new model!!
The bells tolled today, but nobody came to church.