Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Cerebras Code

cerebras.ai

Post reply on HN
Print view View original

141–150 of 185 posts

Re: Cerebras Code

#141

Post by stingraycharles »

echelon wrote: post #97
superasn wrote: post #72

Earlier quoted context omitted.

Pretty sure this is there to prevent this[1] from happening to them [1] https://www.viberank.app/

Oh my god. That's insane. The anti-AI people would be pulling their pitchforks out against these people. Would there be any way of compiling this without people's consent? Looking at GitHub public repos, etc.? I imagine a future where we're all automatically profiled like this. Kind of like perverse employee tracking software.

The pro-AI people are as well, as these people are all on the Claude Max plan, and they’re just burning through resources for internet lols, while ruining the fun for the rest of us. It’s the tragedy of the commons at work.

Re: Cerebras Code

#142

Post by jedisct1 »

I'm a little bit confused.

I subscribed to the $50 plan. It's super fast for sure, but rate limits kick in after just a couple requests. completely defeating the fact that responses are fast.

Did I miss something?

Re: Cerebras Code

#143

Post by saberience »

Ok it's fast, but rate limits seem to kick in extremely quickly and the results are less good than Claude Code and it ends up more expensive?

Who is the intended audience for Cerebras?

Re: Cerebras Code

#145

Post by ixel »

b2m9 wrote: post #137
brandall10 wrote: post #40

Earlier quoted context omitted.

Claude Code does not have a daily limit, it has 5 hour windows that reset. On the $100 plan it's pretty hard to hit a window limit w/ Sonnet unless you're using multiple/subagents. The $200 is better suited to those who do that or want to use a significant amount of Opus. Also the weekly limit selling point is silly - it almost certainly only impacts those who are abusing, ie. running 24/7.

At this point I'm afraid to ask, but I will do I anyways: How do Claude's rate limits actually work? I'm not a Pro/Max5/Max20 subscriber, only light API usage for Anthropic - so it's likely that I don't really understand the limits there. For example, community reports that Anthropic's message limit for Max 5 translates to roughly 88k token per 5-hour window (there's variance, but it's somewhere in this 80-120k ballp…

With Anthropic's Claude subscriptions - while many people appear to use tokens as an idea of the usage limit, I doubt that's what is really used by Anthropic. Why do I say this? Well, there are multiple models, Haiku, Sonnet and Opus, we all know that Opus is the most expensive and burns through the usage limit of the subscription the fastest of all. I'd theorise that Anthropic have some kind of internal credit value (perhaps as simple as $ USD) which they allocate with some variance based on things like overall system load.

Anyway, my personal experience on Max 20x is that, with Opus at least, on a busy day in the past I can burn through between 150 to 200 million tokens in a day using Claude Code for development stuff. Split that up into 5 hour windows, and assume I'm possibly using 2 or 3 windows in a day, that still works out to a lot of tokens, well into the millions. So, the 88k tokens per 5-hour window on Max 5x, I'm not sure if it's really as small as that. Maybe the apparent reductions recently in usage limits have made it drop to around that ballpark. Originally I saw Max 5x as a heavy usage Sonnet plan, with Max 20x being a heavy usage Opus plan, however with the new and additional weekly usage limit happening on August 28th I think I'd see the plans as potentially moderate to heavy usage Sonnet for Max 5x, and heavy usage Sonnet with multiple concurrent agents for Max 20x.

TLDR: I strongly imagine that Claude subscription usage limits are based on some kind of internal credit value, perhaps $ USD, not specifically tokens, and depending which model you use is how fast this "credit" will be depleted.

The usage limits are currently for an account, based on a 5-hour window, from the first message that was sent in a new 5-hour window. From August 28th there's an additional weekly limit which looks like it will primarily make Opus usage restricted.

Re: Cerebras Code

#146

Post by ritenuto »

saberience wrote: post #143

Ok it's fast, but rate limits seem to kick in extremely quickly and the results are less good than Claude Code and it ends up more expensive? Who is the intended audience for Cerebras?

While I’m also curious, I’m fine with having a mostly inferior alternative too. This is a dynamic market with some big players already; having more options is beneficial. If only as a way to prevent others from doing a rug pull.

Re: Cerebras Code

#147

Post by ixel »

The usage limit on Cerebras Code is rather limited, $50 plan apparently gives you 7.5 million tokens per day which doesn't last long. This also isn't clearly advertised on the plans prior to purchasing.

Re: Cerebras Code

#148

Post by diggan »

SamDc73 wrote: post #34
sneilan1 wrote: post #16

Earlier quoted context omitted.

1,000 messages per day should be plenty as a daily development driver. I use claude code sonnet 4 exclusively and I do not send more than 1,000 messages per day. However, that is my current understanding. I am certainly not pressing enter 1,000 times! Maybe there are more messages being sent under the hood that I do not realize?

Still not sure if it's 1000 messages or calls though, if messages that's good.

Neither, it seems. The blog post says "Send up to N messages per day", but the FAQ (https://cerebras-inference.help.usepylon.com/articles/346886...) says:

> How do you calculate messages per day? Actual number of messages per day depends on token usage per request. Estimates based on average requests of ~8k tokens each for a median user.

So seems there is a token limit? But they're not clear what exactly that is? Haven't tried to subscribe, just going by public information available.

Re: Cerebras Code

#149

Post by asaddhamani »

brandall10 wrote: post #40
twothreeone wrote: post #25

Earlier quoted context omitted.

I don't see what's hard to understand about this.. other providers have weekly limits and daily limits. If you max out your daily every day you might still hit your weekly after 3-4 days of usage, meaning you cannot send more for the rest of the week. This is saying that no such weekly limit exists on top of the daily. E.g. see https://techcrunch.com/2025/07/28/anthropic-unveils-new-rate...

Claude Code does not have a daily limit, it has 5 hour windows that reset. On the $100 plan it's pretty hard to hit a window limit w/ Sonnet unless you're using multiple/subagents. The $200 is better suited to those who do that or want to use a significant amount of Opus. Also the weekly limit selling point is silly - it almost certainly only impacts those who are abusing, ie. running 24/7.

Claude now does have a weekly limit so if you are able to hit your weekly (undisclosed, dynamic) limit in 2 days, you're unable to use the services for the next 5 days. That is what Cerebras is referencing with no weekly limits. Claude has session count limits, dynamic limits within each session, and now weekly limits on top of all that.

Re: Cerebras Code

#150

Post by d3vr »

ixel wrote: post #147

The usage limit on Cerebras Code is rather limited, $50 plan apparently gives you 7.5 million tokens per day which doesn't last long. This also isn't clearly advertised on the plans prior to purchasing.

Yeah really disappointing, hopefully they'll reconsider this limit because it really isn't usable, especially with "agentic tools" (e.g: opencode) ..
Post reply on HN