Claude outage - Hacker News

Posted by stuartmemo 10/31/2025

159 points | 186 commentspage 3

miningape 10/31/2025|

Was wondering why we haven't had any fires today - it's nice having some calm on a Friday. Could we make this a weekly thing?

trvz 10/31/2025||

  curl -fsSL https://ollama.com/install.sh | sh
  ollama run smollm2:135m-instruct-q2_K --verbose

brabel 10/31/2025||

Use llama.cpp or llamafile , ollama just adds weight and does too much magic.

tidbeck 10/31/2025||

What kind of tasks do you use it for?

sho 10/31/2025||

5 hours in and I'm pretty disappointed in Anthropic's response to this. Looking at their status page you'd think everything was OK, the downtime was only 37 minutes, a fix is in place and they're "monitoring" - that's total BS. I can't even log in. I can forgive downtime, we've all been there, but I really don't like when companies try to smooth it over with little self-serving minimizations of the disruption.. while it's still in progress!

I use Claude Code for programming work, but I choose OpenAI anywhere customer-facing and this cute little outage is making me feel better about that with every passing minute. NOT cool.

Aeolun 10/31/2025|

Yeah, their API's are returning me a grand total of 1 token per second now or so. It's...

doppp 10/31/2025||

Guys, coding without AI is still a thing.

ukuina 10/31/2025|

Is it really worth the time, though?

You are better off allowing your overworked neural pathways some much-needed rest.

kirpalx 10/31/2025||

Claude AI is down, keeps logging me out.

CuriouslyC 10/31/2025||

Credit for owning downtime, but guys, if people are going to rely on you, you have to figure out solid ops.

roman_soldier 10/31/2025||

Grok seems to have much better uptime, I have Pro sub on both Claude and Grok so just switch between them.

sho 10/31/2025||

Same here. About 2 hours ago it started throwing errors about tool names, now completely down.

roschdal 10/31/2025||

It's a reminder to never rely on something as flaky as AI for your important things.

xdfgh1112 10/31/2025||

Even if AI is down for 20% of the time it can still outcode you 100x.

saagarjha 10/31/2025|||

I’m having trouble getting the AI to outcode me even 1x

erikbye 10/31/2025||

Point me your repo, I want to see if AI can improve it.

saagarjha 11/1/2025|||

I mean I can't give you my company code but my GitHub is not difficult to find. That said, I'm a little confused by your offer, because I feel like this doomed to fail because you don't have my context as to what I want improved?

mirsadm 10/31/2025||||

If the goal is to generate some text then it is probably even more than 100x.

idonotknowwhy 10/31/2025||

You're absolutely right!

miningape 10/31/2025|||

lol, takes like this are why I keep coming back to HN

jstummbillig 10/31/2025|||

AWS being down is not a reminder to not use HTTP or SSH.

askl 10/31/2025||

But it's a reminder to not rely on cloud services.

erikbye 10/31/2025|||

It's a reminder to not rely on a single service.

Logic not strong in this thread.

idonotknowwhy 10/31/2025|||

Yeah lesson learned. But only took 10 minutes to swap out to GLM

https://docs.z.ai/devpack/tool/claude

Smart move making it a drop-in replacement.

paultnylund 10/31/2025|||

[dead]

tooa 10/31/2025||

yepp

Razengan 10/31/2025||

Is it bad form to make feature requests and give feedback on a downtime post? I don't care. I just want this to reach Anthropic's eyes.

Please let us:

* Sign in with Apple on the website

* Buy subscriptions from iOS In App Purchases

* Remove our payment info from the website

* Give paying subscribers an easy way to get actual support

As a frequent traveller I'm not sure if some of those features are gated by region, because some people said they can do some of those things.

remnv 10/31/2025|

Still got issues with the reauth, i’ll try glm for now

More comments...