You can test it now on the official deepseek api. Just set your model to deepseek-v4.1-flash-expires-on-0910

It’s good and very fast.

(Note that the deepseek API trains on your data)

Nah it’s fine. They committed to international devices being unlocked, and the firmware lock is trivial to bypass anyway.

I can believe a lot of people said that. But I'm not sure that means it's a true prediction of consumer behavior.

Easy, Rambo. How about we start with any punishment.

Because he posted about it.

It was an incredibly racist post claiming “gypsies” are like invading wolves and that something more drastic must be done to get rid of them before they kill all the “sheep” in Copenhagen.

It's actually a standard term for the person who plays this role during incident response https://www.pagerduty.com/resources/incident-management-resp...

Hours after they alerted Meta that similar activity was happening from their corporate network?

Read to the end. It isn't really clear how much causality can be inferred from the hours after thing.

Hey if the dad is into tech, decent chance the angry teenage kid is into tech too.

And probably back doors, too. Gotta verify in the downloads though.

Well the porn titles leans it a little.

I get what you're saying and it's concerning how much power these big labs have amassed and how little transparency there is in what they do with it...

But I doubt this a major factor in the trend. I just don't think it's something most corporate users run into. My understanding is these guardrails are negotiable for enterprise customers anyway.

And, not for nothing, but if I owned a human-powered translation company I would've refused to translate it too.

Why? Seems like benchmarks that closely mirror the tasks you'd want an LLM to help with would be a lot more useful than some general intelligence benchmark.

I just did a little anecdotal test. Had pi + cerebras review a recent commit and asked a few quick followups on it. Worked great.

The Cerebras session cost me $1.60 and took a total of 5.1 mins. I did get a few brief 429 rate limit errors in there. The p50 speed was 890 tok/s and 0.64s TTFT.

Using OpenRouter averages, that would've cost $0.29 (no cache discount at Cerebras!) and would've taken about 14.4 minutes.

So on this one short session, cerebras was 5.6x more expensive in exchange for being 2.8x faster. Or, another way, $1.32 buys back about 9 minutes of your time. Not a bad trade IMHO but the cache situation is a real bummer. The longer your session the more relatively expensive Cerebras gets. The "good" news is you're also limited by its short context window.

(Also, I used to be on the Cerebras coding plan and the support is pretty bad for end users. My guess is these public endpoints are really just product demos for potential enterprise customers.)

Thanks! Is there something about their platform that prevents caching? Or are they just not passing on the discount?

The session had a 91.4% cache hit rate. They just give zero discount.

It sounds to me that they don't have enough capacity and they want to discourage people from using the service.

There is nothing about their architecture that prevents reusing the KV cache other than the opportunity cost of keeping the memory occupied.