Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Phyllotaxis: An audio-reactive LED display (jagi.studio)
    12comments
  2. The systems that no one will test (christianperone.com)
    10comments
  3. Booted up in 1993, this server still runs – but not for much longer (2017) (computerworld.com)
    3comments
  4. Pirating the Pirates (mubi.com)
    271comments
  5. California farmers are struggling to sell grapes as demand for wine drops (kqed.org)
    428comments
  6. MicroLLM Lab – Try 7 tiny LLM's in the browser (stateofutopia.com)
    81comments
  7. Tank Body Problem (jimsitu.com)
    21comments
  8. ESP32S3 cluster running 1.58-bit (BitNet) Language model (github.com/low-zi-hong)
    18comments
  9. OpenAI: Tomorrow we are re-opening the Pro $200 subscription (twitter.com/thsottiaux)
    32comments
  10. Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms (github.com/firelex)
    179comments
  11. 12,000-year-old Göbeklitepe burials explain scattered bones (archaeologymag.com)
    30comments
  12. Show HN: Pac-Bench – How well can models one-shot a Pac-Man game? (jonclegg.github.io)
    34comments
  13. Sonnet 5.5 (anthropic.com)
    505comments
  14. Hijacking the PS5's RTMP stream (yashgarg.dev)
    79comments
  15. Simulating Airband AM Radios (bitbashing.io)
    3comments
  16. Scientists solve 1840s space weather mystery (arstechnica.com)
    46comments
  17. Kids turned low-traffic NPR Spotify comments into a secret group chat (thisamericanlife.org)
    214comments
  18. Updated Google Maps shows destruction of the city of Rafah (twitter.com/aliabunimah)
    305comments
  19. Who Killed Paulina Borsook's Career? (wired.com)
    16comments
  20. Uncensored and Offensive Security AI Models Benchmark (github.com/joasasantos)
    —discuss
  21. Nvidia wants to put a watchdog chip next to every AI agent (cnbc.com)
    199comments
  22. World Labs is Joining AMD (worldlabs.ai)
    110comments
  23. 1996 chat room simulator connected to Win95 and System 7 web desktops (lolchat.rip)
    44comments
  24. Show HN: Museum of Numbers (numbermuseum.com)
    8comments
  25. It's Time to Investigate the AI Labs (calnewport.com)
    182comments
  26. How to win a beer with high-dimensional statistics (jamiesimon.io)
    8comments
  27. Show HN: HN.watch – Videos of all Hacker News posts (hn.watch)
    90comments
  28. Does Reddit have an astroturfing problem? What the data suggests (petervijeh.com)
    254comments
  29. What is the best shape of a city? Modelling effect of urban form on distance (sagepub.com)
    20comments
  30. Behold the pawpaw (cbc.ca)
    32comments

OpenAI: Tomorrow we are re-opening the Pro $200 subscription

40 pointsby 1h agotwitter.com
29 comments
1h agoHN ↗

This doesn't bode well, 500$ is the new 200$? Wild that they're doing this after Sonnet/Opus 5.5 releases.

I wonder how legal it is to change usage during a running subscription.

56m agoHN ↗

They also halved API prices with GPT-6, so I think it’ll net out pretty close for sol and Luna, but much less to be used for Astra.

39m agoHN ↗

I wonder how legal it is to change usage during a running subscription.

Legality is no concern for these magical companies. Money is of no object, laws do not apply, the return on investment must not be spoken.

59m agoHN ↗

They're halving the $200 sub so it essentially becomes 10x from 20x

44m agoHN ↗

20x was never about total usage, only 5h one. For 2x price you get 2x usage

32m agoHN ↗

Factually wrong. I mean how can 20x be about 5h usage when 5h limits were not part of the Pro plan for a long time.

30m agoHN ↗

You're wrong for OpenAI, their 5x/20x always referred to the full weekly limit, and my (and others') multiple tests confirmed it empirically multiple times by calculating API-equivalent prices, especially when going from one subscription tier to another.

You're right for Anthropic though, their 5x/20x is mostly for 5h limit.

13m agoHN ↗

Having two basic Pro accounts - one from Anthropic - makes much more sense now.

49m agoHN ↗

Over time, we see prices go low enough that it makes sense for most to buy usage as needed

38m agoHN ↗

Hard to believe when all they've done so far is raise prices and reduce usage quotas.

The 20x plan right now lasts about 1 day and they're reducing it to half a day of usage?

25m agoHN ↗

I agree with them. It will happen circa 2035, when OpenAI will at best be a Microsoft division, acquired for $20bn.

48m agoHN ↗

"it will net out at half the dollar in API spend compared to the old Pro $200 plan"

So effectively 50% less usage.

44m agoHN ↗

we guarantee that over time you always get more work done and with an increasing level of quality

or your money back?

39m agoHN ↗

Codex GPT-6 models are so congested and slow at the moment it is difficult to get through the current 20x plan anyway.

36m agoHN ↗

Well, this is upsetting. I was already maxing out two $200 subscriptions days before the weekly limit. I wonder how bad it'll be after this change.

26m agoHN ↗

So, they’re finally trying to end the subscription subsidies. Presumably ahead of Q4 so that their financials will look better going into next year for the IPO.

23m agoHN ↗

The assumption of the math is that you are not using Astra (which is hilarious). If you are not using Astra there is no point currently when Claude Opus 5.5 exists.

20m agoHN ↗

OpenAI were initially competing with Anthropic in $ value of tokens, and then number of tokens, and then the 5h limit removal meant you could buy multiple plans for huge continuous usage. This gravy train was never sustainable from a margin perspective and must have cost them billions.

It’s now going to be interesting how anthropic react - 5h limit and retain 15-20x $ value, or go the same was as OpenAI.

14m agoHN ↗

Anthropic have the revenue to subsidise longer. OpenAI is lagging behind Anthropic on the API revenue which is where Anthropic is making the majority of its money. I expect Anthropic will smell blood in the water and double down on their subsidies.

10m agoHN ↗

Yes very possibly. If you assume your competition will always exist then I think it makes sense to increase prices as they do even if you have the best business model.

19m agoHN ↗

So after a very tepid response to Sol / Luna 6 - right when opus 5.5 is almost universally praised - OpenAI decided to cut in half their only subscription plan that was still decent value?

Things sure change quickly. 2 months ago you had nigh infinite usage and Sol 5.6 was a significant step forward. Now they're behind in both models and usage, and Astra 6.1 is too dangerous to release or some such (or possibly got embarrassed by opus). What are they even going to show off on today's devday, is a clone of grokbot really that interesting?

15m agoHN ↗

It's very obvious that they need compute for whatever they are announcing tomorrow and that thing will most likely have free usage. So they are cutting usage on the biggest users of compute to allow them to be able to serve whatever they are announcing tomorrow to the masses.

The thing is probably a Muse, and Grok Bot competitor.

14m agoHN ↗

I wish they would give the notice one month in advance. Last week my $200 was renewed, and if tomorrow my usage cuts in half (if I understood correctly) it feels as I got partially ripped.

13m agoHN ↗

Looks like OpenAI is finally compute constrained (and they're already back to the market to raise money, presumably to for more compute).

I'm not sure if compute alone is enough of a moat enough but so far that's how they have survived Anthropic's capture of the enterprise/professional market. Along with shitposting on X and playing Anthropic's unforced errors on developer relations.

But I suspect OpenAI will have to make some hard decisions about de-prioritizing their consumer products, which is most of their current valuation?

8m agoHN ↗

Codex $200 usage was already lower than $200 of Claude for around 2 months now. Half of that means that subscription is not really worth it unless you max things out all the time. But if you're going to go with API rates, there are better priced options out there than whatever OpenAI provides.

This is a risky strategy for OpenAI, and time will tell how will this work. But my hunch is that this is a catastrophic blunder.

I was listening to Dario's podcast where he was talking about how the entire game is in predicting extremely well how much compute you buy in advance, and if you miscalculate, it's pretty much over.

Looks like OpenAI miscalculated.

6m agoHN ↗

People are up in arms about this change but I think the knee-jerk reaction here is missing the bigger picture.

I bet API pricing is going to be more competitive with enterprises.