- Updated: April 6, 2026
- 62 min read
Claude Code Outage Sparks Discussion on Alternatives and Status Updates
Hacker Newsnew | past | comments | ask | show | jobs | submitloginClaude Code Down60 points by theahura 41 minutes ago | hide | past | favorite | 47 commentshttps://www.reddit.com/r/ClaudeCode/comments/1se1p30/broken_again/ https://www.reddit.com/r/ClaudeCode/comments/1se1vn1/oauth_api_key_expiring_daily/ https://www.reddit.com/r/ClaudeCode/comments/1se1ur5/500_error_or_timeout_when_trying_to_reauthorize/Currently nothing on the status: https://status.claude.com/ help user- 5 minutes ago | next [–] I am a believer that everyone should have their main flow be model/provider agnostic at a high level. I often run out of claude tokens and use GLM-5 as backup.https://gist.github.com/ManveerBhullar/7ed5c01a0850d59188632.simple script i use to toggle which backend my claude code is usingreplycyanydeez 2 minutes ago | parent | next [–] Interesting; do you find they actually react the same way to the harness?replyLucasoato 16 minutes ago | prev | next [–] The real issue isn’t that Claude is down, it can happen. The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour. Status pages should be trusted, connected to real metrics and not fake pr stuff :/EDIT: Now they show the issue, kudos to them! Transparency is the key to build trust. No body expects a perfect service, thanks Claude team for your efforts.replychermi 1 minute ago | parent | next [–] This has consistently pissed me off. It seems like we all just accepted that whatever they define as “functioning”/”OK” is suitable. I see the status now shows, but there should be a very loud third party ruthlessly running continuous tests against all of them. Ideally it would also provide proof of the degradation we seem to all agree happens (looking at you Gemini). Like a leaderboard focused on actual live performance.Of course they’d probably quickly game that too. But something showing time to first response, “global capacity reached” etc. The one that pissed me off the lost was Gemini API displaying very clearly 1) user cancelled request in Gemini chat app 2) API showing “user quota reached”. Both were blatant lies. In the latter case, you could find the actual global quota cause later in the error message. I don’t know why there isn’t more outrage.I’m guessing this sort of behavior is not new, but it’s never been so visible to me.replyhn1986 7 minutes ago | parent | prev | next [–] it does show issues. https://status.claude.com/replysimlevesque 4 minutes ago | root | parent | next [–] It shows issues now. Probably not when the person you’re replying to wrote their comment.replyrvz 2 minutes ago | parent | prev | next [–] > The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.When Claude took an extra day off, he forgot to report his hours to the dashboard when he will be unavailable / unresponsive and this is probably why people here are complaining about no status update.Wonder where I have seen that before?replyesafak 12 minutes ago | parent | prev | next [–] You need to use a user-reported status page; the incentives are broken for self reporting.replyinglor_cz 15 minutes ago | parent | prev | next [–] Interesting. I just fixed something using Claude Code. But I am located in Central Europe.replypixl97 13 minutes ago | root | parent | next [–] I’ve not looked into it, but I’d assume they have more than one data center.replyCoolGuySteve 16 minutes ago | prev | next [–] I’m finding qwen 27b is comparable to sonnet but my self hosting has about 5 more 9s than whatever Anthropic’s vibe coding. I also don’t have to worry about the quality of the model I’m being served from day to day.Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are.replyCharlieDigital 3 minutes ago | parent | next [–] > Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are I have been working on some work related to MCP and found some gaps in implementation in Claude and Codex. This is a relatively simple, well-defined spec and both Claude Code and Codex CLI have incomplete/incorrect implementations.During this process of investigation, I checked the CC repo and noticed they had 5000+ issues open.Out of curiosity, I skimmed through them and many point to regressions, real bugs, simple changes, etc. Maybe they have some internal tracker they are using, but you would think that a company with functionally unlimited tokens and access to the best models would be able to use those tokens to get their own house in order.My sense now is that there is a need for the industry to create a lot of hype right now so we see showmanship like the kernel compiler and the agent swarms building a semi-functional browser, etc.yet their own tooling has not fully implemented their own protocol (MCP) correctly.They need all of us to believe that these agents are more capable than they actually are; the more piles of tangled code you write and the more discipline you cede to their LLMs, the more dependent you are on those LLMs to even know what the code is doing. At some point, teams become incapable of teasing the code apart anymore because no one will understand it.Peeking at the issues in the repos and seeing big gaps in functionality like Codex’s missing support for MCP prompts and resources is like looking behind the curtain at reality.replykccqzy 6 minutes ago | parent | prev | next [–] But do you actually treat LLMs as glorified autocomplete or treat them as puzzle solvers where you give them difficult tasks beyond your own intellect?Recently I wrote a data transformation pipeline and I added a note that the whole pipeline should be idempotent.I asked Claude to prove it or find a counterexample. It found one after 25 minutes of thinking; I reasonably estimate that it would take me far longer. I couldn’t care less about using Claude to type code I already knew.replywise0wl 6 minutes ago | parent | prev | next [–] I’ve tried a few models and some are decent, including Qwens models.I’ve tried a few harnesses like Roo Code in VSCode to put things together that in theory emulate the experience I get from VSCode + Claude or Copilot, but I generally find the experience extremely limited and frustrating.How have you set things up to have a good experience?replyCoolGuySteve 3 minutes ago | root | parent | next [–] I’m using the qwen cli tool with a duckduckgo search skill that I made Claude write.It’s like bootstrapping I guessOnce it can search for factual information online the smaller model size becomes less noticeablereplychis 9 minutes ago | parent | prev | next [–] Just to make one obvious critique your costs per token are probably about 1000x higher than the ones they provide.I’m pretty sympathetic to Anthropic/OpenAI just because they are scaling a pretty new technology by 10x every year.It is too bad Google isn’t trying to compete on coding models though, I feel like they’d do way better on the infra and stability side.replytills13 12 minutes ago | parent | prev | next [–] What do you run it on? And even then, I’m guessing your tokens per second are not great?replyCoolGuySteve 12 minutes ago | root | parent | next [–] I get about 35-40tok/sec on a 3090.It’s actually about the same speed when accounting for how much more responsive my system is to Anthropic’s saas infrastructurereplycyanydeez 4 minutes ago | parent | prev | next [–] QWEN3.5-Next-Coder does wonders. It’s drawbacks are time to first token is 30 seconds to load the model and OpenCode has an unsolved timeout issue on this load, but otherwise once it’s warmed up, it’s entirely serviceable.I’ve got a AMD395+ with 128GB, so running a ~46GB model gives me about 85k tokens, which gives me easily copy/paste/find/replace behavior; it mocks up new components; it can wire in some functionality, but that’s usually at it’s limits and requires more debugging.I’ve been looking at how to schedule it using systemd to keep a wiki up to date with a long loaded project and breaks the “blank page” issue with extending behaviors in a side project.I understand some of these larger models can do things faster and smarter, but I don’t see how they can implement novel functionality required for the type of app I’m concerned with. If I just wanted to make endless CRUD or TODO apps, I’m betting I could figure out a loop that’s mostly hands off.replyjasonjmcghee 12 minutes ago | parent | prev | next [–] People keep saying this and idk what I’m doing wrong.Using q8_0 on all the latest and greatest local models and they just don’t come close to sonnet.I’ve tried different harnesses, building my own etc.They are reasonably close to haiku? Maybe?replybakugo 5 minutes ago | root | parent | next [–] You’re not doing anything wrong, they are not comparable.Claims to the tune of “this 0.5B local model running on my phone is almost as good as [large expensive model]” are common but greatly exaggerated, it’s simply not true beyond the most basic use cases.Only the much larger models (such as the 744B GLM-5) manage to come close, but nobody’s running those locally.replytheanonymousone 0 minutes ago | prev | next [–] How much is remaining untilthe last 9 is gone too?replykwertyoowiyop 3 minutes ago | prev | next [–] For older programmers: this is like when Stack Overflow would go down.For really old programmers: this is like when Computer Literacy bookstore was closed.replychermi 7 minutes ago | prev | next [–] Yup displays as an “auth” issue to me. Just a nice reminder that my original plan was to be provider agnostic but everything was working so well with cc I lost sight lol.replysdoering 15 minutes ago | prev | next [–] I am currently – and (this post is up 25 minutes as of now) have been using it without noticeable degradation over the last few hours.Edit: But the status page – at least as of now – is clearly communicating elevated error rates.replyjgrahamc 4 minutes ago | prev | next [–] Here in my corner of Europe it seems to be working fine.replyrishabhaiover 39 minutes ago | prev | next [–] OAuth is failing, I can’t login via claude code.replymatheusmoreira 19 minutes ago | parent | next [–] Same here. Usage limits are still pretty insane too.replybronlund 24 minutes ago | parent | prev | next [–] Same here.replybuildbot 14 minutes ago | prev | next [–] Not that it is the best indicator, but downdector is showing many services with spikes at exactly the same time as Claude Codes issues began.replynightpool 7 minutes ago | parent | next [–] downdector always shows spikes when you go to look at it, and then they remove them later retroactively if the spikes are fake.replyhgoel 19 minutes ago | prev | next [–] Here’s hoping they can get it sorted quick.Hopefully these are just growing pains and not indicative of a GitHub style inability to achieve stability.replyrishabhaiover 2 minutes ago | prev | next [–] The downtime forces me to relook at my utterly dependent relationship with agentic assistance. The inertia to begin engaging with my code is higher than it has ever been.replymatheusmoreira 0 minutes ago | parent | next [–] Yeah. It’s actually starting to make me anxious. I think I got addicted to these agents.replylovecode007 4 minutes ago | prev | next [–] Yup for me too – VSC Claude is def down and not workingreplylovecode007 4 minutes ago | prev | next [–] Def down, keeps saying internal server errorreplykevinwang 24 minutes ago | prev | next [–] Still down for me. (And still nothing on the status page!)replyChrisArchitect 3 minutes ago | prev | next [–] Link for up top: https://status.claude.com/incidents/vfjv5x6qkd4jreplyksajadi 20 minutes ago | prev | next [–] Failures all over Code and Chat here too (London UK) and status is showing all greenreplywojciem 17 minutes ago | prev | next [–] What are decent alternatives to ClaudeCode?replyabi 14 minutes ago | parent | next [–] Codex is great.replyrongenre 12 minutes ago | parent | prev | next [–] I’ve found minimax to be quite goodreplyLeomuck 17 minutes ago | parent | prev | next [–] Probably OpenCode, also works with Claude.replygedy 4 minutes ago | parent | prev | next [–] A keyboard /sreplymistic92 14 minutes ago | prev | next [–] I don’t have any issuesreplyenraged_camel 6 minutes ago | prev | next [–] Claude Code inside the desktop app works for me.replyrvz 8 minutes ago | prev | next [–] Claude isn’t down. He’s on vacation for today and took an extra day off after the weekend.He’ll be back to work by tomorrow.reply Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact Search: [{},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”Claude Code Down60 points by theahura 41 minutes ago | hide | past | favorite | 47 commentshttps://www.reddit.com/r/ClaudeCode/comments/1se1p30/broken_again/\nhttps://www.reddit.com/r/ClaudeCode/comments/1se1vn1/oauth_api_key_expiring_daily/\nhttps://www.reddit.com/r/ClaudeCode/comments/1se1ur5/500_error_or_timeout_when_trying_to_reauthorize/Currently nothing on the status: https://status.claude.com/ help\n\nuser- 5 minutes ago | next [–]\nI am a believer that everyone should have their main flow be model/provider agnostic at a high level. I often run out of claude tokens and use GLM-5 as backup.https://gist.github.com/ManveerBhullar/7ed5c01a0850d59188632.simple script i use to toggle which backend my claude code is usingreplycyanydeez 2 minutes ago | parent | next [–]\nInteresting; do you find they actually react the same way to the harness?replyLucasoato 16 minutes ago | prev | next [–]\nThe real issue isn’t that Claude is down, it can happen. The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.Status pages should be trusted, connected to real metrics and not fake pr stuff :/EDIT: Now they show the issue, kudos to them! Transparency is the key to build trust. No body expects a perfect service, thanks Claude team for your efforts.replychermi 1 minute ago | parent | next [–]\nThis has consistently pissed me off. It seems like we all just accepted that whatever they define as \”functioning\”/\”OK\” is suitable.I see the status now shows, but there should be a very loud third party ruthlessly running continuous tests against all of them. Ideally it would also provide proof of the degradation we seem to all agree happens (looking at you Gemini). Like a leaderboard focused on actual live performance. Of course they’d probably quickly game that too. But something showing time to first response, \”global capacity reached\” etc.The one that pissed me off the lost was Gemini API displaying very clearly 1) user cancelled request in Gemini chat app 2) API showing \”user quota reached\”. Both were blatant lies. In the latter case, you could find the actual global quota cause later in the error message. I don’t know why there isn’t more outrage. I’m guessing this sort of behavior is not new, but it’s never been so visible to me.replyhn1986 7 minutes ago | parent | prev | next [–]\nit does show issues. \nhttps://status.claude.com/replysimlevesque 4 minutes ago | root | parent | next [–]\nIt shows issues now. Probably not when the person you’re replying to wrote their comment.replyrvz 2 minutes ago | parent | prev | next [–]\n> The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.When Claude took an extra day off, he forgot to report his hours to the dashboard when he will be unavailable / unresponsive and this is probably why people here are complaining about no status update.Wonder where I have seen that before?replyesafak 12 minutes ago | parent | prev | next [–]\nYou need to use a user-reported status page; the incentives are broken for self reporting.replyinglor_cz 15 minutes ago | parent | prev | next [–]\nInteresting. I just fixed something using Claude Code.But I am located in Central Europe.replypixl97 13 minutes ago | root | parent | next [–]\nI’ve not looked into it, but I’d assume they have more than one data center.replyCoolGuySteve 16 minutes ago | prev | next [–]\nI’m finding qwen 27b is comparable to sonnet but my self hosting has about 5 more 9s than whatever Anthropic’s vibe coding. I also don’t have to worry about the quality of the model I’m being served from day to day.Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are.replyCharlieDigital 3 minutes ago | parent | next [–]\n > Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are\n\nI have been working on some work related to MCP and found some gaps in implementation in Claude and Codex.This is a relatively simple, well-defined spec and both Claude Code and Codex CLI have incomplete/incorrect implementations.During this process of investigation, I checked the CC repo and noticed they had 5000+ issues open. Out of curiosity, I skimmed through them and many point to regressions, real bugs, simple changes, etc.Maybe they have some internal tracker they are using, but you would think that a company with functionally unlimited tokens and access to the best models would be able to use those tokens to get their own house in order.My sense now is that there is a need for the industry to create a lot of hype right now so we see showmanship like the kernel compiler and the agent swarms building a semi-functional browser, etc.yet their own tooling has not fully implemented their own protocol (MCP) correctly.They need all of us to believe that these agents are more capable than they actually are; the more piles of tangled code you write and the more discipline you cede to their LLMs, the more dependent you are on those LLMs to even know what the code is doing. At some point, teams become incapable of teasing the code apart anymore because no one will understand it.Peeking at the issues in the repos and seeing big gaps in functionality like Codex’s missing support for MCP prompts and resources is like looking behind the curtain at reality.replykccqzy 6 minutes ago | parent | prev | next [–]\nBut do you actually treat LLMs as glorified autocomplete or treat them as puzzle solvers where you give them difficult tasks beyond your own intellect?Recently I wrote a data transformation pipeline and I added a note that the whole pipeline should be idempotent.I asked Claude to prove it or find a counterexample. It found one after 25 minutes of thinking; I reasonably estimate that it would take me far longer. I couldn’t care less about using Claude to type code I already knew.replywise0wl 6 minutes ago | parent | prev | next [–]\nI’ve tried a few models and some are decent, including Qwens models.I’ve tried a few harnesses like Roo Code in VSCode to put things together that in theory emulate the experience I get from VSCode + Claude or Copilot, but I generally find the experience extremely limited and frustrating.How have you set things up to have a good experience?replyCoolGuySteve 3 minutes ago | root | parent | next [–]\nI’m using the qwen cli tool with a duckduckgo search skill that I made Claude write.It’s like bootstrapping I guessOnce it can search for factual information online the smaller model size becomes less noticeablereplychis 9 minutes ago | parent | prev | next [–]\nJust to make one obvious critique your costs per token are probably about 1000x higher than the ones they provide.I’m pretty sympathetic to Anthropic/OpenAI just because they are scaling a pretty new technology by 10x every year.It is too bad Google isn’t trying to compete on coding models though, I feel like they’d do way better on the infra and stability side.replytills13 12 minutes ago | parent | prev | next [–]\nWhat do you run it on? And even then, I’m guessing your tokens per second are not great?replyCoolGuySteve 12 minutes ago | root | parent | next [–]\nI get about 35-40tok/sec on a 3090.It’s actually about the same speed when accounting for how much more responsive my system is to Anthropic’s saas infrastructurereplycyanydeez 4 minutes ago | parent | prev | next [–]\nQWEN3.5-Next-Coder does wonders. It’s drawbacks are time to first token is 30 seconds to load the model and OpenCode has an unsolved timeout issue on this load, but otherwise once it’s warmed up, it’s entirely serviceable.I’ve got a AMD395+ with 128GB, so running a ~46GB model gives me about 85k tokens, which gives me easily copy/paste/find/replace behavior; it mocks up new components; it can wire in some functionality, but that’s usually at it’s limits and requires more debugging.I’ve been looking at how to schedule it using systemd to keep a wiki up to date with a long loaded project and breaks the \”blank page\” issue with extending behaviors in a side project.I understand some of these larger models can do things faster and smarter, but I don’t see how they can implement novel functionality required for the type of app I’m concerned with. If I just wanted to make endless CRUD or TODO apps, I’m betting I could figure out a loop that’s mostly hands off.replyjasonjmcghee 12 minutes ago | parent | prev | next [–]\nPeople keep saying this and idk what I’m doing wrong.Using q8_0 on all the latest and greatest local models and they just don’t come close to sonnet.I’ve tried different harnesses, building my own etc.They are reasonably close to haiku? Maybe?replybakugo 5 minutes ago | root | parent | next [–]\nYou’re not doing anything wrong, they are not comparable.Claims to the tune of \”this 0.5B local model running on my phone is almost as good as [large expensive model]\” are common but greatly exaggerated, it’s simply not true beyond the most basic use cases.Only the much larger models (such as the 744B GLM-5) manage to come close, but nobody’s running those locally.replytheanonymousone 0 minutes ago | prev | next [–]\nHow much is remaining untilthe last 9 is gone too?replykwertyoowiyop 3 minutes ago | prev | next [–]\nFor older programmers: this is like when Stack Overflow would go down.For really old programmers: this is like when Computer Literacy bookstore was closed.replychermi 7 minutes ago | prev | next [–]\nYup displays as an \”auth\” issue to me. Just a nice reminder that my original plan was to be provider agnostic but everything was working so well with cc I lost sight lol.replysdoering 15 minutes ago | prev | next [–]\nI am currently – and (this post is up 25 minutes as of now) have been using it without noticeable degradation over the last few hours.Edit: But the status page – at least as of now – is clearly communicating elevated error rates.replyjgrahamc 4 minutes ago | prev | next [–]\nHere in my corner of Europe it seems to be working fine.replyrishabhaiover 39 minutes ago | prev | next [–]\nOAuth is failing, I can’t login via claude code.replymatheusmoreira 19 minutes ago | parent | next [–]\nSame here. Usage limits are still pretty insane too.replybronlund 24 minutes ago | parent | prev | next [–]\nSame here.replybuildbot 14 minutes ago | prev | next [–]\nNot that it is the best indicator, but downdector is showing many services with spikes at exactly the same time as Claude Codes issues began.replynightpool 7 minutes ago | parent | next [–]\ndowndector always shows spikes when you go to look at it, and then they remove them later retroactively if the spikes are fake.replyhgoel 19 minutes ago | prev | next [–]\nHere’s hoping they can get it sorted quick. Hopefully these are just growing pains and not indicative of a GitHub style inability to achieve stability.replyrishabhaiover 2 minutes ago | prev | next [–]\nThe downtime forces me to relook at my utterly dependent relationship with agentic assistance. The inertia to begin engaging with my code is higher than it has ever been.replymatheusmoreira 0 minutes ago | parent | next [–]\nYeah. It’s actually starting to make me anxious. I think I got addicted to these agents.replylovecode007 4 minutes ago | prev | next [–]\nYup for me too – VSC Claude is def down and not workingreplylovecode007 4 minutes ago | prev | next [–]\nDef down, keeps saying internal server errorreplykevinwang 24 minutes ago | prev | next [–]\nStill down for me. (And still nothing on the status page!)replyChrisArchitect 3 minutes ago | prev | next [–]\nLink for up top: https://status.claude.com/incidents/vfjv5x6qkd4jreplyksajadi 20 minutes ago | prev | next [–]\nFailures all over Code and Chat here too (London UK) and status is showing all greenreplywojciem 17 minutes ago | prev | next [–]\nWhat are decent alternatives to ClaudeCode?replyabi 14 minutes ago | parent | next [–]\nCodex is great.replyrongenre 12 minutes ago | parent | prev | next [–]\nI’ve found minimax to be quite goodreplyLeomuck 17 minutes ago | parent | prev | next [–]\nProbably OpenCode, also works with Claude.replygedy 4 minutes ago | parent | prev | next [–]\nA keyboard /sreplymistic92 14 minutes ago | prev | next [–]\nI don’t have any issuesreplyenraged_camel 6 minutes ago | prev | next [–]\nClaude Code inside the desktop app works for me.replyrvz 8 minutes ago | prev | next [–]\nClaude isn’t down. He’s on vacation for today and took an extra day off after the weekend.He’ll be back to work by tomorrow.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”Claude Code Down”,”user- 5 minutes ago | next [–]\nI am a believer that everyone should have their main flow be model/provider agnostic at a high level. I often run out of claude tokens and use GLM-5 as backup.https://gist.github.com/ManveerBhullar/7ed5c01a0850d59188632.simple script i use to toggle which backend my claude code is usingreply”:”60 points by theahura 41 minutes ago | hide | past | favorite | 47 comments”,”cyanydeez 2 minutes ago | parent | next [–]\nInteresting; do you find they actually react the same way to the harness?reply”:””,”Lucasoato 16 minutes ago | prev | next [–]\nThe real issue isn’t that Claude is down, it can happen.The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour. Status pages should be trusted, connected to real metrics and not fake pr stuff :/EDIT: Now they show the issue, kudos to them! Transparency is the key to build trust. No body expects a perfect service, thanks Claude team for your efforts.reply”:”https://www.reddit.com/r/ClaudeCode/comments/1se1p30/broken_again/\nhttps://www.reddit.com/r/ClaudeCode/comments/1se1vn1/oauth_api_key_expiring_daily/\nhttps://www.reddit.com/r/ClaudeCode/comments/1se1ur5/500_error_or_timeout_when_trying_to_reauthorize/Currently nothing on the status: https://status.claude.com/”,”chermi 1 minute ago | parent | next [–]\nThis has consistently pissed me off. It seems like we all just accepted that whatever they define as \”functioning\”/\”OK\” is suitable.I see the status now shows, but there should be a very loud third party ruthlessly running continuous tests against all of them. Ideally it would also provide proof of the degradation we seem to all agree happens (looking at you Gemini). Like a leaderboard focused on actual live performance. Of course they’d probably quickly game that too. But something showing time to first response, \”global capacity reached\” etc.The one that pissed me off the lost was Gemini API displaying very clearly 1) user cancelled request in Gemini chat app 2) API showing \”user quota reached\”. Both were blatant lies. In the latter case, you could find the actual global quota cause later in the error message. I don’t know why there isn’t more outrage. I’m guessing this sort of behavior is not new, but it’s never been so visible to me.reply”:””,”hn1986 7 minutes ago | parent | prev | next [–]\nit does show issues.\nhttps://status.claude.com/reply”:”help”,”simlevesque 4 minutes ago | root | parent | next [–]\nIt shows issues now. Probably not when the person you’re replying to wrote their comment.reply”:”user- 5 minutes ago | next [–]\nI am a believer that everyone should have their main flow be model/provider agnostic at a high level. I often run out of claude tokens and use GLM-5 as backup.https://gist.github.com/ManveerBhullar/7ed5c01a0850d59188632.simple script i use to toggle which backend my claude code is usingreply”,”rvz 2 minutes ago | parent | prev | next [–]\n> The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.When Claude took an extra day off, he forgot to report his hours to the dashboard when he will be unavailable / unresponsive and this is probably why people here are complaining about no status update.Wonder where I have seen that before?reply”:””,”esafak 12 minutes ago | parent | prev | next [–]\nYou need to use a user-reported status page; the incentives are broken for self reporting.reply”:””,”inglor_cz 15 minutes ago | parent | prev | next [–]\nInteresting. I just fixed something using Claude Code. But I am located in Central Europe.reply”:”user- 5 minutes ago | next [–]\nI am a believer that everyone should have their main flow be model/provider agnostic at a high level.I often run out of claude tokens and use GLM-5 as backup.https://gist.github.com/ManveerBhullar/7ed5c01a0850d59188632.simple script i use to toggle which backend my claude code is usingreply”,”pixl97 13 minutes ago | root | parent | next [–]\nI’ve not looked into it, but I’d assume they have more than one data center.reply”:”cyanydeez 2 minutes ago | parent | next [–]\nInteresting; do you find they actually react the same way to the harness?reply”,”CoolGuySteve 16 minutes ago | prev | next [–]\nI’m finding qwen 27b is comparable to sonnet but my self hosting has about 5 more 9s than whatever Anthropic’s vibe coding. I also don’t have to worry about the quality of the model I’m being served from day to day.Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are.reply”:””,”CharlieDigital 3 minutes ago | parent | next [–]\n > Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are\n\nI have been working on some work related to MCP and found some gaps in implementation in Claude and Codex. This is a relatively simple, well-defined spec and both Claude Code and Codex CLI have incomplete/incorrect implementations.During this process of investigation, I checked the CC repo and noticed they had 5000+ issues open. Out of curiosity, I skimmed through them and many point to regressions, real bugs, simple changes, etc. Maybe they have some internal tracker they are using, but you would think that a company with functionally unlimited tokens and access to the best models would be able to use those tokens to get their own house in order.My sense now is that there is a need for the industry to create a lot of hype right now so we see showmanship like the kernel compiler and the agent swarms building a semi-functional browser, etc.yet their own tooling has not fully implemented their own protocol (MCP) correctly.They need all of us to believe that these agents are more capable than they actually are; the more piles of tangled code you write and the more discipline you cede to their LLMs, the more dependent you are on those LLMs to even know what the code is doing. At some point, teams become incapable of teasing the code apart anymore because no one will understand it.Peeking at the issues in the repos and seeing big gaps in functionality like Codex’s missing support for MCP prompts and resources is like looking behind the curtain at reality.reply”:””,”kccqzy 6 minutes ago | parent | prev | next [–]\nBut do you actually treat LLMs as glorified autocomplete or treat them as puzzle solvers where you give them difficult tasks beyond your own intellect?Recently I wrote a data transformation pipeline and I added a note that the whole pipeline should be idempotent.I asked Claude to prove it or find a counterexample. It found one after 25 minutes of thinking; I reasonably estimate that it would take me far longer. I couldn’t care less about using Claude to type code I already knew.reply”:”cyanydeez 2 minutes ago | parent | next [–]\nInteresting; do you find they actually react the same way to the harness?reply”,”wise0wl 6 minutes ago | parent | prev | next [–]\nI’ve tried a few models and some are decent, including Qwens models.I’ve tried a few harnesses like Roo Code in VSCode to put things together that in theory emulate the experience I get from VSCode + Claude or Copilot, but I generally find the experience extremely limited and frustrating.How have you set things up to have a good experience?reply”:”Lucasoato 16 minutes ago | prev | next [–]\nThe real issue isn’t that Claude is down, it can happen.The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour. Status pages should be trusted, connected to real metrics and not fake pr stuff :/EDIT: Now they show the issue, kudos to them! Transparency is the key to build trust. No body expects a perfect service, thanks Claude team for your efforts.reply”,”CoolGuySteve 3 minutes ago | root | parent | next [–]\nI’m using the qwen cli tool with a duckduckgo search skill that I made Claude write. It’s like bootstrapping I guessOnce it can search for factual information online the smaller model size becomes less noticeablereply”:””,”chis 9 minutes ago | parent | prev | next [–]\nJust to make one obvious critique your costs per token are probably about 1000x higher than the ones they provide.I’m pretty sympathetic to Anthropic/OpenAI just because they are scaling a pretty new technology by 10x every year. It is too bad Google isn’t trying to compete on coding models though, I feel like they’d do way better on the infra and stability side.reply”:””,”tills13 12 minutes ago | parent | prev | next [–]\nWhat do you run it on? And even then, I’m guessing your tokens per second are not great?reply”:”Lucasoato 16 minutes ago | prev | next [–]\nThe real issue isn’t that Claude is down, it can happen. The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour. Status pages should be trusted, connected to real metrics and not fake pr stuff :/EDIT: Now they show the issue, kudos to them! Transparency is the key to build trust. No body expects a perfect service, thanks Claude team for your efforts.reply”,”CoolGuySteve 12 minutes ago | root | parent | next [–]\nI get about 35-40tok/sec on a 3090.It’s actually about the same speed when accounting for how much more responsive my system is to Anthropic’s saas infrastructurereply”:”chermi 1 minute ago | parent | next [–]\nThis has consistently pissed me off. It seems like we all just accepted that whatever they define as \”functioning\”/\”OK\” is suitable.I see the status now shows, but there should be a very loud third party ruthlessly running continuous tests against all of them. Ideally it would also provide proof of the degradation we seem to all agree happens (looking at you Gemini). Like a leaderboard focused on actual live performance. Of course they’d probably quickly game that too. But something showing time to first response, \”global capacity reached\” etc.The one that pissed me off the lost was Gemini API displaying very clearly 1) user cancelled request in Gemini chat app 2) API showing \”user quota reached\”. Both were blatant lies. In the latter case, you could find the actual global quota cause later in the error message. I don’t know why there isn’t more outrage. I’m guessing this sort of behavior is not new, but it’s never been so visible to me.reply”,”cyanydeez 4 minutes ago | parent | prev | next [–]\nQWEN3.5-Next-Coder does wonders.It’s drawbacks are time to first token is 30 seconds to load the model and OpenCode has an unsolved timeout issue on this load, but otherwise once it’s warmed up, it’s entirely serviceable.I’ve got a AMD395+ with 128GB, so running a ~46GB model gives me about 85k tokens, which gives me easily copy/paste/find/replace behavior; it mocks up new components; it can wire in some functionality, but that’s usually at it’s limits and requires more debugging.I’ve been looking at how to schedule it using systemd to keep a wiki up to date with a long loaded project and breaks the \”blank page\” issue with extending behaviors in a side project.I understand some of these larger models can do things faster and smarter, but I don’t see how they can implement novel functionality required for the type of app I’m concerned with. If I just wanted to make endless CRUD or TODO apps, I’m betting I could figure out a loop that’s mostly hands off.reply”:””,”jasonjmcghee 12 minutes ago | parent | prev | next [–]\nPeople keep saying this and idk what I’m doing wrong. Using q8_0 on all the latest and greatest local models and they just don’t come close to sonnet.I’ve tried different harnesses, building my own etc.They are reasonably close to haiku? Maybe?reply”:””,”bakugo 5 minutes ago | root | parent | next [–]\nYou’re not doing anything wrong, they are not comparable.Claims to the tune of \”this 0.5B local model running on my phone is almost as good as [large expensive model]\” are common but greatly exaggerated, it’s simply not true beyond the most basic use cases.Only the much larger models (such as the 744B GLM-5) manage to come close, but nobody’s running those locally.reply”:”chermi 1 minute ago | parent | next [–]\nThis has consistently pissed me off. It seems like we all just accepted that whatever they define as \”functioning\”/\”OK\” is suitable.I see the status now shows, but there should be a very loud third party ruthlessly running continuous tests against all of them. Ideally it would also provide proof of the degradation we seem to all agree happens (looking at you Gemini). Like a leaderboard focused on actual live performance. Of course they’d probably quickly game that too. But something showing time to first response, \”global capacity reached\” etc.The one that pissed me off the lost was Gemini API displaying very clearly 1) user cancelled request in Gemini chat app 2) API showing \”user quota reached\”. Both were blatant lies. In the latter case, you could find the actual global quota cause later in the error message. I don’t know why there isn’t more outrage. I’m guessing this sort of behavior is not new, but it’s never been so visible to me.reply”,”theanonymousone 0 minutes ago | prev | next [–]\nHow much is remaining untilthe last 9 is gone too?reply”:”hn1986 7 minutes ago | parent | prev | next [–]\nit does show issues. \nhttps://status.claude.com/reply”,”kwertyoowiyop 3 minutes ago | prev | next [–]\nFor older programmers: this is like when Stack Overflow would go down.For really old programmers: this is like when Computer Literacy bookstore was closed.reply”:””,”chermi 7 minutes ago | prev | next [–]\nYup displays as an \”auth\” issue to me. Just a nice reminder that my original plan was to be provider agnostic but everything was working so well with cc I lost sight lol.reply”:””,”sdoering 15 minutes ago | prev | next [–]\nI am currently – and (this post is up 25 minutes as of now) have been using it without noticeable degradation over the last few hours.Edit: But the status page – at least as of now – is clearly communicating elevated error rates.reply”:”hn1986 7 minutes ago | parent | prev | next [–]\nit does show issues. \nhttps://status.claude.com/reply”,”jgrahamc 4 minutes ago | prev | next [–]\nHere in my corner of Europe it seems to be working fine.reply”:”simlevesque 4 minutes ago | root | parent | next [–]\nIt shows issues now. Probably not when the person you’re replying to wrote their comment.reply”,”rishabhaiover 39 minutes ago | prev | next [–]\nOAuth is failing, I can’t login via claude code.reply”:””,”matheusmoreira 19 minutes ago | parent | next [–]\nSame here. Usage limits are still pretty insane too.reply”:””,”bronlund 24 minutes ago | parent | prev | next [–]\nSame here.reply”:”simlevesque 4 minutes ago | root | parent | next [–]\nIt shows issues now. Probably not when the person you’re replying to wrote their comment.reply”,”buildbot 14 minutes ago | prev | next [–]\nNot that it is the best indicator, but downdector is showing many services with spikes at exactly the same time as Claude Codes issues began.reply”:”rvz 2 minutes ago | parent | prev | next [–]\n> The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.When Claude took an extra day off, he forgot to report his hours to the dashboard when he will be unavailable / unresponsive and this is probably why people here are complaining about no status update.Wonder where I have seen that before?reply”,”nightpool 7 minutes ago | parent | next [–]\ndowndector always shows spikes when you go to look at it, and then they remove them later retroactively if the spikes are fake.reply”:””,”hgoel 19 minutes ago | prev | next [–]\nHere’s hoping they can get it sorted quick. Hopefully these are just growing pains and not indicative of a GitHub style inability to achieve stability.reply”:””,”rishabhaiover 2 minutes ago | prev | next [–]\nThe downtime forces me to relook at my utterly dependent relationship with agentic assistance. The inertia to begin engaging with my code is higher than it has ever been.reply”:”rvz 2 minutes ago | parent | prev | next [–]\n> The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.When Claude took an extra day off, he forgot to report his hours to the dashboard when he will be unavailable / unresponsive and this is probably why people here are complaining about no status update.Wonder where I have seen that before?reply”,”matheusmoreira 0 minutes ago | parent | next [–]\nYeah.It’s actually starting to make me anxious. I think I got addicted to these agents.reply”:”esafak 12 minutes ago | parent | prev | next [–]\nYou need to use a user-reported status page; the incentives are broken for self reporting.reply”,”lovecode007 4 minutes ago | prev | next [–]\nYup for me too – VSC Claude is def down and not workingreply”:””,”lovecode007 4 minutes ago | prev | next [–]\nDef down, keeps saying internal server errorreply”:””,”kevinwang 24 minutes ago | prev | next [–]\nStill down for me. (And still nothing on the status page!)reply”:”esafak 12 minutes ago | parent | prev | next [–]\nYou need to use a user-reported status page; the incentives are broken for self reporting.reply”,”ChrisArchitect 3 minutes ago | prev | next [–]\nLink for up top: https://status.claude.com/incidents/vfjv5x6qkd4jreply”:”inglor_cz 15 minutes ago | parent | prev | next [–]\nInteresting. I just fixed something using Claude Code. But I am located in Central Europe.reply”,”ksajadi 20 minutes ago | prev | next [–]\nFailures all over Code and Chat here too (London UK) and status is showing all greenreply”:””,”wojciem 17 minutes ago | prev | next [–]\nWhat are decent alternatives to ClaudeCode?reply”:””,”abi 14 minutes ago | parent | next [–]\nCodex is great.reply”:”inglor_cz 15 minutes ago | parent | prev | next [–]\nInteresting. I just fixed something using Claude Code. But I am located in Central Europe.reply”,”rongenre 12 minutes ago | parent | prev | next [–]\nI’ve found minimax to be quite goodreply”:”pixl97 13 minutes ago | root | parent | next [–]\nI’ve not looked into it, but I’d assume they have more than one data center.reply”,”Leomuck 17 minutes ago | parent | prev | next [–]\nProbably OpenCode, also works with Claude.reply”:””,”gedy 4 minutes ago | parent | prev | next [–]\nA keyboard /sreply”:””,”mistic92 14 minutes ago | prev | next [–]\nI don’t have any issuesreply”:”pixl97 13 minutes ago | root | parent | next [–]\nI’ve not looked into it, but I’d assume they have more than one data center.reply”,”enraged_camel 6 minutes ago | prev | next [–]\nClaude Code inside the desktop app works for me.reply”:”CoolGuySteve 16 minutes ago | prev | next [–]\nI’m finding qwen 27b is comparable to sonnet but my self hosting has about 5 more 9s than whatever Anthropic’s vibe coding. I also don’t have to worry about the quality of the model I’m being served from day to day.Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are.reply”,”rvz 8 minutes ago | prev | next [–]\nClaude isn’t down.He’s on vacation for today and took an extra day off after the weekend.He’ll be back to work by tomorrow.reply”:””},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:””,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:”60 points by theahura 41 minutes ago | hide | past | favorite | 47 comments”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:””,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:”https://www.reddit.com/r/ClaudeCode/comments/1se1p30/broken_again/\nhttps://www.reddit.com/r/ClaudeCode/comments/1se1vn1/oauth_api_key_expiring_daily/\nhttps://www.reddit.com/r/ClaudeCode/comments/1se1ur5/500_error_or_timeout_when_trying_to_reauthorize/Currently nothing on the status: https://status.claude.com/”},{},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:””,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:”help”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”cyanydeez 2 minutes ago | parent | next [–]\nInteresting; do you find they actually react the same way to the harness?reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”cyanydeez 2 minutes ago | parent | next [–]\nInteresting; do you find they actually react the same way to the harness?reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”Lucasoato 16 minutes ago | prev | next [–]\nThe real issue isn’t that Claude is down, it can happen.The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour. Status pages should be trusted, connected to real metrics and not fake pr stuff :/EDIT: Now they show the issue, kudos to them! Transparency is the key to build trust. No body expects a perfect service, thanks Claude team for your efforts.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”Lucasoato 16 minutes ago | prev | next [–]\nThe real issue isn’t that Claude is down, it can happen. The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour. Status pages should be trusted, connected to real metrics and not fake pr stuff :/EDIT: Now they show the issue, kudos to them! Transparency is the key to build trust.No body expects a perfect service, thanks Claude team for your efforts.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”chermi 1 minute ago | parent | next [–]\nThis has consistently pissed me off. It seems like we all just accepted that whatever they define as \”functioning\”/\”OK\” is suitable. I see the status now shows, but there should be a very loud third party ruthlessly running continuous tests against all of them.Ideally it would also provide proof of the degradation we seem to all agree happens (looking at you Gemini). Like a leaderboard focused on actual live performance. Of course they’d probably quickly game that too. But something showing time to first response, \”global capacity reached\” etc. The one that pissed me off the lost was Gemini API displaying very clearly 1) user cancelled request in Gemini chat app 2) API showing \”user quota reached\”. Both were blatant lies.In the latter case, you could find the actual global quota cause later in the error message. I don’t know why there isn’t more outrage. I’m guessing this sort of behavior is not new, but it’s never been so visible to me.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”chermi 1 minute ago | parent | next [–]\nThis has consistently pissed me off.It seems like we all just accepted that whatever they define as \”functioning\”/\”OK\” is suitable. I see the status now shows, but there should be a very loud third party ruthlessly running continuous tests against all of them. Ideally it would also provide proof of the degradation we seem to all agree happens (looking at you Gemini). Like a leaderboard focused on actual live performance. Of course they’d probably quickly game that too.But something showing time to first response, \”global capacity reached\” etc. The one that pissed me off the lost was Gemini API displaying very clearly 1) user cancelled request in Gemini chat app 2) API showing \”user quota reached\”. Both were blatant lies. In the latter case, you could find the actual global quota cause later in the error message. I don’t know why there isn’t more outrage. I’m guessing this sort of behavior is not new, but it’s never been so visible to me.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”hn1986 7 minutes ago | parent | prev | next [–]\nit does show issues. \nhttps://status.claude.com/reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”hn1986 7 minutes ago | parent | prev | next [–]\nit does show issues. \nhttps://status.claude.com/reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”simlevesque 4 minutes ago | root | parent | next [–]\nIt shows issues now. Probably not when the person you’re replying to wrote their comment.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”simlevesque 4 minutes ago | root | parent | next [–]\nIt shows issues now. Probably not when the person you’re replying to wrote their comment.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”rvz 2 minutes ago | parent | prev | next [–]\n> The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.When Claude took an extra day off, he forgot to report his hours to the dashboard when he will be unavailable / unresponsive and this is probably why people here are complaining about no status update.Wonder where I have seen that before?reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”rvz 2 minutes ago | parent | prev | next [–]\n> The problem is that the status page doesn’t report anything, even if it has been impossible to log in during the past hour.When Claude took an extra day off, he forgot to report his hours to the dashboard when he will be unavailable / unresponsive and this is probably why people here are complaining about no status update.Wonder where I have seen that before?reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”esafak 12 minutes ago | parent | prev | next [–]\nYou need to use a user-reported status page; the incentives are broken for self reporting.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”esafak 12 minutes ago | parent | prev | next [–]\nYou need to use a user-reported status page; the incentives are broken for self reporting.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”inglor_cz 15 minutes ago | parent | prev | next [–]\nInteresting. I just fixed something using Claude Code. But I am located in Central Europe.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”inglor_cz 15 minutes ago | parent | prev | next [–]\nInteresting. I just fixed something using Claude Code. But I am located in Central Europe.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”pixl97 13 minutes ago | root | parent | next [–]\nI’ve not looked into it, but I’d assume they have more than one data center.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”pixl97 13 minutes ago | root | parent | next [–]\nI’ve not looked into it, but I’d assume they have more than one data center.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”CoolGuySteve 16 minutes ago | prev | next [–]\nI’m finding qwen 27b is comparable to sonnet but my self hosting has about 5 more 9s than whatever Anthropic’s vibe coding. I also don’t have to worry about the quality of the model I’m being served from day to day.Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”CoolGuySteve 16 minutes ago | prev | next [–]\nI’m finding qwen 27b is comparable to sonnet but my self hosting has about 5 more 9s than whatever Anthropic’s vibe coding. I also don’t have to worry about the quality of the model I’m being served from day to day.Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”CharlieDigital 3 minutes ago | parent | next [–]\n > Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are\n\nI have been working on some work related to MCP and found some gaps in implementation in Claude and Codex. This is a relatively simple, well-defined spec and both Claude Code and Codex CLI have incomplete/incorrect implementations.During this process of investigation, I checked the CC repo and noticed they had 5000+ issues open. Out of curiosity, I skimmed through them and many point to regressions, real bugs, simple changes, etc. Maybe they have some internal tracker they are using, but you would think that a company with functionally unlimited tokens and access to the best models would be able to use those tokens to get their own house in order.My sense now is that there is a need for the industry to create a lot of hype right now so we see showmanship like the kernel compiler and the agent swarms building a semi-functional browser, etc.yet their own tooling has not fully implemented their own protocol (MCP) correctly.They need all of us to believe that these agents are more capable than they actually are; the more piles of tangled code you write and the more discipline you cede to their LLMs, the more dependent you are on those LLMs to even know what the code is doing. At some point, teams become incapable of teasing the code apart anymore because no one will understand it.Peeking at the issues in the repos and seeing big gaps in functionality like Codex’s missing support for MCP prompts and resources is like looking behind the curtain at reality.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”CharlieDigital 3 minutes ago | parent | next [–]\n > Probably the most damning fact about LLMs is just how poorly written their parent companies’ systems are\n\nI have been working on some work related to MCP and found some gaps in implementation in Claude and Codex. This is a relatively simple, well-defined spec and both Claude Code and Codex CLI have incomplete/incorrect implementations.During this process of investigation, I checked the CC repo and noticed they had 5000+ issues open. Out of curiosity, I skimmed through them and many point to regressions, real bugs, simple changes, etc. Maybe they have some internal tracker they are using, but you would think that a company with functionally unlimited tokens and access to the best models would be able to use those tokens to get their own house in order.My sense now is that there is a need for the industry to create a lot of hype right now so we see showmanship like the kernel compiler and the agent swarms building a semi-functional browser, etc.yet their own tooling has not fully implemented their own protocol (MCP) correctly.They need all of us to believe that these agents are more capable than they actually are; the more piles of tangled code you write and the more discipline you cede to their LLMs, the more dependent you are on those LLMs to even know what the code is doing. At some point, teams become incapable of teasing the code apart anymore because no one will understand it.Peeking at the issues in the repos and seeing big gaps in functionality like Codex’s missing support for MCP prompts and resources is like looking behind the curtain at reality.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”kccqzy 6 minutes ago | parent | prev | next [–]\nBut do you actually treat LLMs as glorified autocomplete or treat them as puzzle solvers where you give them difficult tasks beyond your own intellect?Recently I wrote a data transformation pipeline and I added a note that the whole pipeline should be idempotent. I asked Claude to prove it or find a counterexample. It found one after 25 minutes of thinking; I reasonably estimate that it would take me far longer. I couldn’t care less about using Claude to type code I already knew.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”kccqzy 6 minutes ago | parent | prev | next [–]\nBut do you actually treat LLMs as glorified autocomplete or treat them as puzzle solvers where you give them difficult tasks beyond your own intellect?Recently I wrote a data transformation pipeline and I added a note that the whole pipeline should be idempotent. I asked Claude to prove it or find a counterexample.It found one after 25 minutes of thinking; I reasonably estimate that it would take me far longer. I couldn’t care less about using Claude to type code I already knew.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”wise0wl 6 minutes ago | parent | prev | next [–]\nI’ve tried a few models and some are decent, including Qwens models.I’ve tried a few harnesses like Roo Code in VSCode to put things together that in theory emulate the experience I get from VSCode + Claude or Copilot, but I generally find the experience extremely limited and frustrating.How have you set things up to have a good experience?reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”wise0wl 6 minutes ago | parent | prev | next [–]\nI’ve tried a few models and some are decent, including Qwens models.I’ve tried a few harnesses like Roo Code in VSCode to put things together that in theory emulate the experience I get from VSCode + Claude or Copilot, but I generally find the experience extremely limited and frustrating.How have you set things up to have a good experience?reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”CoolGuySteve 3 minutes ago | root | parent | next [–]\nI’m using the qwen cli tool with a duckduckgo search skill that I made Claude write.It’s like bootstrapping I guessOnce it can search for factual information online the smaller model size becomes less noticeablereply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”CoolGuySteve 3 minutes ago | root | parent | next [–]\nI’m using the qwen cli tool with a duckduckgo search skill that I made Claude write.It’s like bootstrapping I guessOnce it can search for factual information online the smaller model size becomes less noticeablereply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”chis 9 minutes ago | parent | prev | next [–]\nJust to make one obvious critique your costs per token are probably about 1000x higher than the ones they provide.I’m pretty sympathetic to Anthropic/OpenAI just because they are scaling a pretty new technology by 10x every year.It is too bad Google isn’t trying to compete on coding models though, I feel like they’d do way better on the infra and stability side.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”chis 9 minutes ago | parent | prev | next [–]\nJust to make one obvious critique your costs per token are probably about 1000x higher than the ones they provide.I’m pretty sympathetic to Anthropic/OpenAI just because they are scaling a pretty new technology by 10x every year. It is too bad Google isn’t trying to compete on coding models though, I feel like they’d do way better on the infra and stability side.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”tills13 12 minutes ago | parent | prev | next [–]\nWhat do you run it on? And even then, I’m guessing your tokens per second are not great?reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”tills13 12 minutes ago | parent | prev | next [–]\nWhat do you run it on? And even then, I’m guessing your tokens per second are not great?reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”CoolGuySteve 12 minutes ago | root | parent | next [–]\nI get about 35-40tok/sec on a 3090.It’s actually about the same speed when accounting for how much more responsive my system is to Anthropic’s saas infrastructurereply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”CoolGuySteve 12 minutes ago | root | parent | next [–]\nI get about 35-40tok/sec on a 3090.It’s actually about the same speed when accounting for how much more responsive my system is to Anthropic’s saas infrastructurereply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”cyanydeez 4 minutes ago | parent | prev | next [–]\nQWEN3.5-Next-Coder does wonders. It’s drawbacks are time to first token is 30 seconds to load the model and OpenCode has an unsolved timeout issue on this load, but otherwise once it’s warmed up, it’s entirely serviceable.I’ve got a AMD395+ with 128GB, so running a ~46GB model gives me about 85k tokens, which gives me easily copy/paste/find/replace behavior; it mocks up new components; it can wire in some functionality, but that’s usually at it’s limits and requires more debugging.I’ve been looking at how to schedule it using systemd to keep a wiki up to date with a long loaded project and breaks the \”blank page\” issue with extending behaviors in a side project.I understand some of these larger models can do things faster and smarter, but I don’t see how they can implement novel functionality required for the type of app I’m concerned with. If I just wanted to make endless CRUD or TODO apps, I’m betting I could figure out a loop that’s mostly hands off.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”cyanydeez 4 minutes ago | parent | prev | next [–]\nQWEN3.5-Next-Coder does wonders.It’s drawbacks are time to first token is 30 seconds to load the model and OpenCode has an unsolved timeout issue on this load, but otherwise once it’s warmed up, it’s entirely serviceable.I’ve got a AMD395+ with 128GB, so running a ~46GB model gives me about 85k tokens, which gives me easily copy/paste/find/replace behavior; it mocks up new components; it can wire in some functionality, but that’s usually at it’s limits and requires more debugging.I’ve been looking at how to schedule it using systemd to keep a wiki up to date with a long loaded project and breaks the \”blank page\” issue with extending behaviors in a side project.I understand some of these larger models can do things faster and smarter, but I don’t see how they can implement novel functionality required for the type of app I’m concerned with. If I just wanted to make endless CRUD or TODO apps, I’m betting I could figure out a loop that’s mostly hands off.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”jasonjmcghee 12 minutes ago | parent | prev | next [–]\nPeople keep saying this and idk what I’m doing wrong. Using q8_0 on all the latest and greatest local models and they just don’t come close to sonnet.I’ve tried different harnesses, building my own etc.They are reasonably close to haiku? Maybe?reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”jasonjmcghee 12 minutes ago | parent | prev | next [–]\nPeople keep saying this and idk what I’m doing wrong. Using q8_0 on all the latest and greatest local models and they just don’t come close to sonnet.I’ve tried different harnesses, building my own etc.They are reasonably close to haiku? Maybe?reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”bakugo 5 minutes ago | root | parent | next [–]\nYou’re not doing anything wrong, they are not comparable.Claims to the tune of \”this 0.5B local model running on my phone is almost as good as [large expensive model]\” are common but greatly exaggerated, it’s simply not true beyond the most basic use cases.Only the much larger models (such as the 744B GLM-5) manage to come close, but nobody’s running those locally.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”bakugo 5 minutes ago | root | parent | next [–]\nYou’re not doing anything wrong, they are not comparable.Claims to the tune of \”this 0.5B local model running on my phone is almost as good as [large expensive model]\” are common but greatly exaggerated, it’s simply not true beyond the most basic use cases.Only the much larger models (such as the 744B GLM-5) manage to come close, but nobody’s running those locally.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”theanonymousone 0 minutes ago | prev | next [–]\nHow much is remaining untilthe last 9 is gone too?reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”theanonymousone 0 minutes ago | prev | next [–]\nHow much is remaining untilthe last 9 is gone too?reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”kwertyoowiyop 3 minutes ago | prev | next [–]\nFor older programmers: this is like when Stack Overflow would go down.For really old programmers: this is like when Computer Literacy bookstore was closed.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”kwertyoowiyop 3 minutes ago | prev | next [–]\nFor older programmers: this is like when Stack Overflow would go down.For really old programmers: this is like when Computer Literacy bookstore was closed.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”chermi 7 minutes ago | prev | next [–]\nYup displays as an \”auth\” issue to me.Just a nice reminder that my original plan was to be provider agnostic but everything was working so well with cc I lost sight lol.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”chermi 7 minutes ago | prev | next [–]\nYup displays as an \”auth\” issue to me. Just a nice reminder that my original plan was to be provider agnostic but everything was working so well with cc I lost sight lol.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”sdoering 15 minutes ago | prev | next [–]\nI am currently – and (this post is up 25 minutes as of now) have been using it without noticeable degradation over the last few hours.Edit: But the status page – at least as of now – is clearly communicating elevated error rates.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”sdoering 15 minutes ago | prev | next [–]\nI am currently – and (this post is up 25 minutes as of now) have been using it without noticeable degradation over the last few hours.Edit: But the status page – at least as of now – is clearly communicating elevated error rates.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”jgrahamc 4 minutes ago | prev | next [–]\nHere in my corner of Europe it seems to be working fine.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”jgrahamc 4 minutes ago | prev | next [–]\nHere in my corner of Europe it seems to be working fine.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”rishabhaiover 39 minutes ago | prev | next [–]\nOAuth is failing, I can’t login via claude code.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”rishabhaiover 39 minutes ago | prev | next [–]\nOAuth is failing, I can’t login via claude code.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”matheusmoreira 19 minutes ago | parent | next [–]\nSame here. Usage limits are still pretty insane too.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”matheusmoreira 19 minutes ago | parent | next [–]\nSame here. Usage limits are still pretty insane too.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”bronlund 24 minutes ago | parent | prev | next [–]\nSame here.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”bronlund 24 minutes ago | parent | prev | next [–]\nSame here.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”buildbot 14 minutes ago | prev | next [–]\nNot that it is the best indicator, but downdector is showing many services with spikes at exactly the same time as Claude Codes issues began.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”buildbot 14 minutes ago | prev | next [–]\nNot that it is the best indicator, but downdector is showing many services with spikes at exactly the same time as Claude Codes issues began.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”nightpool 7 minutes ago | parent | next [–]\ndowndector always shows spikes when you go to look at it, and then they remove them later retroactively if the spikes are fake.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”nightpool 7 minutes ago | parent | next [–]\ndowndector always shows spikes when you go to look at it, and then they remove them later retroactively if the spikes are fake.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”hgoel 19 minutes ago | prev | next [–]\nHere’s hoping they can get it sorted quick.Hopefully these are just growing pains and not indicative of a GitHub style inability to achieve stability.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”hgoel 19 minutes ago | prev | next [–]\nHere’s hoping they can get it sorted quick. Hopefully these are just growing pains and not indicative of a GitHub style inability to achieve stability.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”rishabhaiover 2 minutes ago | prev | next [–]\nThe downtime forces me to relook at my utterly dependent relationship with agentic assistance. The inertia to begin engaging with my code is higher than it has ever been.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”rishabhaiover 2 minutes ago | prev | next [–]\nThe downtime forces me to relook at my utterly dependent relationship with agentic assistance. The inertia to begin engaging with my code is higher than it has ever been.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”matheusmoreira 0 minutes ago | parent | next [–]\nYeah. It’s actually starting to make me anxious.I think I got addicted to these agents.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”matheusmoreira 0 minutes ago | parent | next [–]\nYeah. It’s actually starting to make me anxious. I think I got addicted to these agents.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”lovecode007 4 minutes ago | prev | next [–]\nYup for me too – VSC Claude is def down and not workingreply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”lovecode007 4 minutes ago | prev | next [–]\nYup for me too – VSC Claude is def down and not workingreply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”lovecode007 4 minutes ago | prev | next [–]\nDef down, keeps saying internal server errorreply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”lovecode007 4 minutes ago | prev | next [–]\nDef down, keeps saying internal server errorreply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”kevinwang 24 minutes ago | prev | next [–]\nStill down for me. (And still nothing on the status page!)reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”kevinwang 24 minutes ago | prev | next [–]\nStill down for me. (And still nothing on the status page!)reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”ChrisArchitect 3 minutes ago | prev | next [–]\nLink for up top: https://status.claude.com/incidents/vfjv5x6qkd4jreply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”ChrisArchitect 3 minutes ago | prev | next [–]\nLink for up top: https://status.claude.com/incidents/vfjv5x6qkd4jreply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”ksajadi 20 minutes ago | prev | next [–]\nFailures all over Code and Chat here too (London UK) and status is showing all greenreply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”ksajadi 20 minutes ago | prev | next [–]\nFailures all over Code and Chat here too (London UK) and status is showing all greenreply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”wojciem 17 minutes ago | prev | next [–]\nWhat are decent alternatives to ClaudeCode?reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”wojciem 17 minutes ago | prev | next [–]\nWhat are decent alternatives to ClaudeCode?reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”abi 14 minutes ago | parent | next [–]\nCodex is great.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”abi 14 minutes ago | parent | next [–]\nCodex is great.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”rongenre 12 minutes ago | parent | prev | next [–]\nI’ve found minimax to be quite goodreply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”rongenre 12 minutes ago | parent | prev | next [–]\nI’ve found minimax to be quite goodreply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”Leomuck 17 minutes ago | parent | prev | next [–]\nProbably OpenCode, also works with Claude.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”Leomuck 17 minutes ago | parent | prev | next [–]\nProbably OpenCode, also works with Claude.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”gedy 4 minutes ago | parent | prev | next [–]\nA keyboard /sreply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”gedy 4 minutes ago | parent | prev | next [–]\nA keyboard /sreply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”mistic92 14 minutes ago | prev | next [–]\nI don’t have any issuesreply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”mistic92 14 minutes ago | prev | next [–]\nI don’t have any issuesreply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”enraged_camel 6 minutes ago | prev | next [–]\nClaude Code inside the desktop app works for me.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”enraged_camel 6 minutes ago | prev | next [–]\nClaude Code inside the desktop app works for me.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”rvz 8 minutes ago | prev | next [–]\nClaude isn’t down. He’s on vacation for today and took an extra day off after the weekend.He’ll be back to work by tomorrow.reply”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””,”login”:””,”Claude Code Down”:”rvz 8 minutes ago | prev | next [–]\nClaude isn’t down. He’s on vacation for today and took an extra day off after the weekend.He’ll be back to work by tomorrow.reply”},{“Hacker Newsnew | past | comments | ask | show | jobs | submitlogin”:”Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact\nSearch:”,”Hacker Newsnew | past | comments | ask | show | jobs | submit”:””}]