Interesting tests being done but I can't help but think it limits testing innovation in some way given that the requested apps are essentially all clones of others
It's really interesting to see the Sol/Terra/Luna apps side-by-side.
I need to add these stats somewhere in the UI, but one interesting take away: Terra took 1/2 as much wall-clock time as Sol, but Luna took more wall-clock time than Sol (by about 23%). It's still much much cheaper, but it seems like Terra is likely a more optimal time/cost balance for most use cases.
The Terra quality is usually nearly as good as Sol, but much faster and cheaper. I do appreciate Sol's design sensibilities (see, for example, the audio sequencer). It's the first model in a while that is clearly distinct on that front. They'd all converged to very similar visuals for a while.
"This isn't objective." Correct, and we are not pretending it is. We are not handing down a scientific verdict.
Actually, you are doing rational investigation in a fuzzy probabilistic new/emergent space, with open sharing to the world. I don’t understand why people downplay themselves and put on a pedestal others supposedly serious sciences.
It's a preemptive defense against methodology cynicism seen often on sites including but not limited to Hacker News. I've been guilty of including such defenses myself over the years because I've gotten annoyed with receiving such cynicism.
Ugh, that is bad, and I'm sorry I didn't see it at the top of the thread yesterday.
Really, the deeper problem is the upvotes that cause such posts to rise to the top of a thread and stick there, drowning out curious conversation and giving people a bad impression of the entire community. Unfortunately, the upvote problem seems basically unsolvable - people don't do it consciously and it's very much a tragedy-of-the-commons problem. So we're stuck with moderation on the comments.
because if you don’t put this disclaimer the top comment is always "Acthually this isn't real science because you didn't publish your P value" so you can't win.
also the article itself is clearly LLM generated though
Because for its entire existence, the top HN comment on articles is typically a contrarian take or pointing out flaws. This goes double for a study, where people just hunt for some aspect of the methodology they dislike. If you don't address the flaws, then it looks like you never considered them, and the top comment will say that your entire methodology is suspect. It's super predictable to the point that you can harness this kind of reaction to get stuff on the frontpage if you really want to.
Ultimately advocates exist for models and there are incredible financial incentives for some to be advocates, so authors are guaranteed someone being mad if their horse doesn't perform well.
Given that type of reaction is inevitable, it just saves the conversation.
> We generated a big pile of artifacts, we are publishing all of them, and you can form your own opinion.
My opinion is that two gimmicky "one-shot prompting shootout" marketing pieces in two days smells like desperation. I'm not sure you understand what a turnoff this is for potential customers.
Say you were interviewing a human, to see how capable they were. You are allowed to give them take home work. What kind of questions would you ask, or tasks would you give, to try to get a measure of their competence? If you gave them a task, would you iterate with them on the design, or would you see what they could produce on their own, without input?
Measuring "intelligence" is hard, but giving an "intelligent" entity tasks, and seeing what comes out, and then comparing the output with others, seems like a very reasonable, relative, way to do it.
> Even if this was a good idea when applied to humans (it's not)
I'm not sure I understand. What's not a good idea? I'm asking you how you would do it, with some possible examples. Or, are you saying it's a bad idea to try to measure how competent someone is before hiring them?
> LLMs aren't humans,
Not sure how this is relevant. My question was how to measure competence and "intelligence" for a task a human needs done. LLMs are, currently, used to complete tasks humans want completed that would usually be done by humans. That's why there's so much money being dumped into token usage. I, personally, have trouble considering tasks outside of what humans can do. What do you have in mind?
There's a ton of arenamaxxing going on (especially from facebook), though I don't disagree that it's one of the better actual benchmarks.
Always fun to ask them to recreate classic demoscene effects (sadly they're still pretty bad at generating music, though at least claude seems to create decent synths).
I keep trying to get them to recreate the fluid+particle stuff from Agenda Circling Forth etc., but even giving them the blog posts describing the implementation (and screenshots) they're still pretty bad.
Arena can definitely be benchmaxxed a bit if you tried to. The distribution of prompts there is very different than usage by regular coders. E.g., lots of requests for one-shot games from scratch.
My concern with most of these visual benchmarks, popular as they are, is that they are likely more indicative of knowledge (i.e. how comprehensive the training data is and how well it can be retrieved from the model) than of reasoning ability. I don't see in particular how a model would construct a CoT that mapped somehow to a representation of the cube geometry and its animations in latent space without a large chunk of that being pre-existing information.
> without a large chunk of that being pre-existing information.
Is there any evidence that novel reasoning is present in LLM? I've never been able to make that work, and I believe Apple's paper some time ago was good evidence that it doesn't exist. In my experience, sparse latent spaces result in a complete, comical, failure in reasoning.
It’s interesting how all the model names and versions are like SKUS taking up space on a display shelf. I look forward to whatever Sagittarius A* does!
> Separate question, separate table. This is our standard latency harness (three short prompts, five reps, 400-token cap), not the build tasks. tok/s is output tokens over wall-clock, uniform for all.
> so their tok/s is a ceiling, not a true decode rate. The clear read: the GPT-5.6 tiers are the snappiest models here on short prompts (Luna answers in about a second), Qwen is absurdly cheap and fast, and DeepSeek and GLM are the slowpokes
You put in a lot of good work, and kudos for that, but man, reading paragraphs like these just puts me off of the entire piece.
Like…how hard would it have been really to type these two sentences by hand, in your own natural voice?
You can suggest a different one and it works pretty well, the bar is very low.
I asked for hemingway in a planning document one time and the result was highly amusing to everyone. "We will not wreck it with small greeds." was an all time favorite for me.
Please write like a normal human and put the effort in to type what you want to say. Using AI to make your writing is not only lazy, it's bland, tiresome, and disrespectful of the reader's time.
Is this how I learn that Bezos now has a beard? Interesting that it is a detail that all of the models chose to include (unless that was in the prompt and just not put in the post).
Obviously AI-written, but I'm confused with the results: Muse Spark has the best Rubik's cube by far, the only one properly animating, yet it gets a 2/5
Really nice breakdown, surprised by the results - especially the fact that OSS models were so behind on most task... (lol at the SVG of the moon without any sign of life by GLM-5.2)
Missing the exact prompts - would love to replicate...but also curious how you prompted these: they could be a big reason why some models failed completely at rendering SVGs (ie. GLM 5.2)
Yeah, the models have all been really good at generating greenfield apps for a really long time (in the scope of LLM time).
I suppose it’s interesting to see how they make better greenfield apps. But I am much more interested in how they solve hard problems in existing gnarly codebases.
One-shot benchmarks is great for me as an solo creator, since it slightly correlates to whether the better frontier models (Opus and Fable for me) make better decisions about things I didn't spec, or whether they'll give me better suggestions right off the bat.
I imagine one could one-shot a basic app and then feed feature requests one by one, sounds like an obvious way to benchmark architecture/maintainability
A lot of these are visual-heavy tests that often require first person sight to confirm results. Considering GLM isn’t multimodal, that might explain why it did better on the calculator question and not much else.
I think there's approximately zero value in seeing how a model can turn 100 tokens into a 100k. What workflow is that? It's not useful in the real world.
I want to know how well it can follow instructions, manage various potentially competing desires in the context, and so on. It's much more interesting how it can turn 100k tokens (e.g. a codebase and lots of tool calls) into 100 tokens.
73 comments
[ 2.8 ms ] story [ 32.1 ms ] threadhttps://arena.logic.inc/
It's really interesting to see the Sol/Terra/Luna apps side-by-side.
I need to add these stats somewhere in the UI, but one interesting take away: Terra took 1/2 as much wall-clock time as Sol, but Luna took more wall-clock time than Sol (by about 23%). It's still much much cheaper, but it seems like Terra is likely a more optimal time/cost balance for most use cases.
The Terra quality is usually nearly as good as Sol, but much faster and cheaper. I do appreciate Sol's design sensibilities (see, for example, the audio sequencer). It's the first model in a while that is clearly distinct on that front. They'd all converged to very similar visuals for a while.
Look at the top comment on their previous HN submission: https://news.ycombinator.com/item?id=48839886
Really, the deeper problem is the upvotes that cause such posts to rise to the top of a thread and stick there, drowning out curious conversation and giving people a bad impression of the entire community. Unfortunately, the upvote problem seems basically unsolvable - people don't do it consciously and it's very much a tragedy-of-the-commons problem. So we're stuck with moderation on the comments.
also the article itself is clearly LLM generated though
Given that type of reaction is inevitable, it just saves the conversation.
https://www.tryai.dev/models/grok-4.5
＊onanizing…
Nothing hard. Everybody wins.
My opinion is that two gimmicky "one-shot prompting shootout" marketing pieces in two days smells like desperation. I'm not sure you understand what a turnoff this is for potential customers.
Measuring "intelligence" is hard, but giving an "intelligent" entity tasks, and seeing what comes out, and then comparing the output with others, seems like a very reasonable, relative, way to do it.
Even if this was a good idea when applied to humans (it's not), LLMs aren't humans, and I worry about people who don't understand the difference.
I'm not sure I understand. What's not a good idea? I'm asking you how you would do it, with some possible examples. Or, are you saying it's a bad idea to try to measure how competent someone is before hiring them?
> LLMs aren't humans,
Not sure how this is relevant. My question was how to measure competence and "intelligence" for a task a human needs done. LLMs are, currently, used to complete tasks humans want completed that would usually be done by humans. That's why there's so much money being dumped into token usage. I, personally, have trouble considering tasks outside of what humans can do. What do you have in mind?
Agent: https://arena.ai/leaderboard/agent
Web dev: https://arena.ai/leaderboard/code/webdev
Currently Fable and 5.6 are neck and neck
Always fun to ask them to recreate classic demoscene effects (sadly they're still pretty bad at generating music, though at least claude seems to create decent synths).
I keep trying to get them to recreate the fluid+particle stuff from Agenda Circling Forth etc., but even giving them the blog posts describing the implementation (and screenshots) they're still pretty bad.
Is there any evidence that novel reasoning is present in LLM? I've never been able to make that work, and I believe Apple's paper some time ago was good evidence that it doesn't exist. In my experience, sparse latent spaces result in a complete, comical, failure in reasoning.
I’m not very valiant to verify its veracity. But even if the math is merely derivative it merits mention.
> so their tok/s is a ceiling, not a true decode rate. The clear read: the GPT-5.6 tiers are the snappiest models here on short prompts (Luna answers in about a second), Qwen is absurdly cheap and fast, and DeepSeek and GLM are the slowpokes
You put in a lot of good work, and kudos for that, but man, reading paragraphs like these just puts me off of the entire piece.
Like…how hard would it have been really to type these two sentences by hand, in your own natural voice?
On the other hand, do we have to complain about every seemingly AI written text?
I asked for hemingway in a planning document one time and the result was highly amusing to everyone. "We will not wreck it with small greeds." was an all time favorite for me.
Man generates text.
Their writing was already painted into this corner long before the LLM epoch and they continue to publish more than anyone else.
Please write like a normal human and put the effort in to type what you want to say. Using AI to make your writing is not only lazy, it's bland, tiresome, and disrespectful of the reader's time.
I'm sick and tired of reading comments like these.
(edit: seems to be an issue with inline videos)
https://d1md4c6gq9re9p.cloudfront.net/blog/gpt-5.6-buildoff/...
https://d1md4c6gq9re9p.cloudfront.net/blog/gpt-5.6-buildoff/...
https://d1md4c6gq9re9p.cloudfront.net/blog/gpt-5.6-buildoff/...
I suppose it’s interesting to see how they make better greenfield apps. But I am much more interested in how they solve hard problems in existing gnarly codebases.
surprised it isn't a bigger thing, eg artificial analysis doesn't report anything like that
still doesn't measure the human-agent interaction part, but that's pure vibes atp
> Draw a horse riding an astronaut in svg
https://www.svgviewer.dev/s/if4gi3e7
I want to know how well it can follow instructions, manage various potentially competing desires in the context, and so on. It's much more interesting how it can turn 100k tokens (e.g. a codebase and lots of tool calls) into 100 tokens.
We made Grok 4.5, GPT-5.5, and Claude build the same apps - https://news.ycombinator.com/item?id=48838772 - July 2026 (92 comments)
Real world is messy, other benchmarks are clearly gameable by the Chinese open models.
Great job! And I don’t care about the tone of the article, it’s readable just fine.
The lessons that should have been learned here, surely, include:
1) you probably should not one-shot apps like this unless you're really not that bothered with consistency
2) if you are remaining in control of the code you generate, Qwen 3.7 plus is pretty competitive with Fable.
My questions:
How is "good results when it worked" a 4/5 score?
And how can any of these really be considered indicators of performance on the "genuinely novel" when the results are all so similar?