17 comments

  • freakynit 4 minutes ago
    It one-shotted generation of Java bindings for this project: https://github.com/jeffhajewski/latticedb

    Related PR: https://github.com/jeffhajewski/latticedb/pull/5

    The session used ~100K input tokens, ~60K output tokens, and ~80K thinking tokens.

    I reviewed it using gpt-sol-medium, and it seems to be satisfied with it's work.

  • giamma 44 minutes ago
    • KellyCriterion 4 minutes ago
      thanks for pointing me out on Unwall.App!

      Didnt know they exist - looks very good, maybe even better than Archive.ph

  • harlan_pdx 19 minutes ago
    Releasing weights is the right move. Keeps them competitive with DeepSeek on the open side.
    • stanac 4 minutes ago
      I had good experience with GLM 5.3, but...

      Z.AI is the only provider for GLM 5.3 on OpenRouter. I don't see 5.3 on Hugging Face. Not sure if this new model is "full GLM" or something smaller, or if they will like Moonshot AI publish weights but put restrictive license [1], which will again leave Z.AI as single GLM model provider on OpenRouter.

      [1] https://huggingface.co/moonshotai/Kimi-K3/blob/main/LICENSE

  • WithinReason 59 minutes ago
    Mixed signals, here it's performing below even GPT-5.4 Nano:

    https://livebench.ai/

    while here it outperforms Fable by a significant margin:

    https://oxalpha.com/

    but if the latter is true, will people still say it was "distilled" from Fable?

    • woadwarrior01 24 minutes ago
      That benchmark is super sus. Until someone pointed it out, the top performing open weights model was a Kimi K3 fine tune from their sponsor (abacusai/Smaug-Agentic). Now, it's not on the list.

      Source: https://twitterwebviewer.com/?tweet=2091116504787935350

    • sunbum 55 minutes ago
      the 2nd website is not official, just something someone slopped together for some reason.
      • yorwba 25 minutes ago
        Even if it weren't slopped together, 65% vs 80% on 10 tasks just isn't a significant difference. For 80% power to distinguish at a significance level of 0.05, you'd need more like 140 samples, if those were the true success probabilities.

        The number one problem in LLM benchmarking is that people try to draw conclusions from sample sizes far too small to conclude anything but "it works sometimes, it fails sometimes, hard to say which is better." (The number two problem is that people run benchmarks blindly without checking that they measure something meaningful.)

      • Alifatisk 40 minutes ago
        I have plenty of these websites, I can’t understand why someone is doing this.
        • colesantiago 16 minutes ago
          It is called phishing and grifting.

          Many people and even software engineers fall for this all the time.

          Most of these people are from crypto pivoting to AI doing this.

          AI has made this easier and cheaper and it is going to get a LOT worse.

          Imagine lots of websites with typosquatting and looking exactly the same as another website, vibe coded and cloned within seconds.

          The public have no chance.

    • tescreal 16 minutes ago
      I really want to see hard evidence of distillation before I buy into it. Seems like a lot of sour grapes over not having the sort of lead assumed. In this field, it has been shown repeatedly that leaps in performance come swiftly and without notice.
    • re-thc 38 minutes ago
      the outperform Fable was a mid (not completed) benchmark run. Real results were lower.
    • epolanski 49 minutes ago
      GLM 5.3 was a great model, so this would be strange to release a regressed model
      • ImprobableTruth 44 minutes ago
        It's probably GLM 5.3 flash, so weaker but cheaper.
        • re-thc 37 minutes ago
          With vision on top
  • esskay 1 hour ago
    I'd be interested to know what was going on with it during the public test as there were numerous reports of it improving considerably at tasks it was asked to do early on in the test compared to later in it.
    • utilize1808 28 minutes ago
      It's logical to serve the best version (quant) of the model at the beginning so that users keep testing it. It is also reasonable to think that the developer of the model tried to test various quant levels by gradually degrading the model's capabilities.
    • daveyoung 57 minutes ago
      Two potentials from my pov:

      1. Just variance in pass@K. If you prompt any model multiple times you'll see a large variance. N=1, but I find chinese open source models have a higher variance than higher-RL'd models like fable/opus.

      2. They legitimately shipped a new RL checkpoint over the 7 days, which I find hard to believe.

      I am leaning towards 1.

      • zarzavat 37 minutes ago
        3. Deployment problems unrelated to the weights causing degraded performance
      • re-thc 37 minutes ago
        2. There was a new checkpoint. Official.
    • rfoo 58 minutes ago
      lol don't shout out the obvious
  • seydor 29 minutes ago
    Funny how all china companies are expected to release weights by default
    • respectattentio 26 minutes ago
      they are playing a completely different game than the US
  • fen_wick 8 minutes ago
    Good to see more competition in the open weights space. The more players the better.
  • glimshe 15 minutes ago
    There's a lot of brand confusion among the Chinese models right now. Kimi, Qwen, GLM, Z.ai, Ox. We might know the difference (or I should say, someone does because I'm losing track already) but these models have no chance at end user penetration and loyalty until there's a single focused survivor.

    It took me a year talking about it until my wife knew that ChatGPT and Gemini are two different things.

    • giwook 11 minutes ago
      I disagree. I think most developers who are savvy enough to be using openweight models and/or running models locally are not dealing with the same level of confusion you are.

      Ox is just GLM. And z.ai is the maker of GLM.

      The main players in the openweight model market have been known for a while.

      And they already have significant user penetration.

    • marclove 5 minutes ago
      Consumers aren’t the customer.
    • tokai 5 minutes ago
      Just because you're confused doesn't mean that there is general confusion here. Its really not that complicated.
  • garo-pro 1 hour ago
    Unfortunately I can't find sources other than this for now but this seems to be legit.
    • mohsen1 43 minutes ago
      > The company on Wednesday confirmed speculation that the Ox Alpha model is a new iteration of its GLM series and said it will release the weights for it tonight, in response to queries by Bloomberg News.

      Seems legit.

      It's really hard to know how good it is. So much hype around it.

    • KaseyKim 29 minutes ago
      they have confirmed it officially
  • j_maffe 1 hour ago
    Anyone has a link to a report of its capabilities? I can't find a reliable source.
    • vblanco 1 hour ago
      completely vibes based, but ive been using it to port Mindustry game from Java to C# with agents, and its been working for 50 hours (its 15-20 tks so super slow inference). Its done a fantastic work and its almost finished now. Better results than deepseek flash and gpt luna by a mile on this kind of long term work. Less good than gpt sol or opus. We dont know the param count but my guess is 200-300 range.
    • daveyoung 54 minutes ago
      likely a distilled glm 5.3 that will punch within 20% of that at 2-3x less size. you'll find that capability is typically very jagged on models that are distilled
  • tosh 53 minutes ago
    my guess is this is a small model punching way above its weight

    on toy benches it made quite a few mistakes but was able to fix all of them on its own

    (meaning more tokens, more turns, more tool calls — but same outcome as gpt 5.6 sol)

  • respectattentio 25 minutes ago
    it's for sure better than deepseek flash 07/31
  • dgellow 1 hour ago
    Do we know the size of the model?
  • kosolam 14 minutes ago
    Only reason people are interested is it’s free at the moment. I wasn’t impressed by its performance. Once the model gets a price tag it’s usage will be negligible.
    • kosolam 12 minutes ago
      It doesn’t rival deepseek v4 flash, and of course not deepseek v4 pro. This is my own impression.
  • daveyoung 45 minutes ago
    [dead]
  • hncsiocp9x 1 hour ago
    [dead]