BEGIN TRAN;
-- put the query here
commit;
I feel like I haven’t had to prod a model to actually do what I told it to in awhile so that was a shock. I guess that it does use fewer tokens that way, just annoying when I’m paying for the “cutting edge” model to have it be lazy on me like that.
This is in Cursor the model popped up and so I tried it out from the model selector.
If so, I’m not sure I’d even consider that a problem. If the goal is for it to give you a query to run, and you ask it “let’s do it in a transaction”, it’s a reasonable thing for it to simply inform you, “yeah you can just type begin first” since it’s assuming you’re going to be pasting the query in anyway. And yeah, it does use fewer tokens, assuming the query was long. Similar to how, if it gave me a command to run, and I say “I’m getting a permission denied”, it would be reasonable for it to say “yeah do it as root, put sudo before the command”, and it’s IMO reasonable if it didn’t repeat the whole thing verbatim just with the word “sudo” first.
But if the context was that you actually expected it to run the query for you, and instead it just said “here, you run it”, then yeah that’s lazy and I’d understand the shock.
All earnesty aside, OpenAI’s oddly specific singular focus on “intelligence per token” (also in the benchmarks) that literally noone else pushes so hard eerily reminds me of Apple’s Macbook anorexia era pre-M1. One metric to chase at the cost of literally anything else. GPT-5.3+ are some of the smartest models out there and could be a pleasure to work with, if they weren’t lazy bastards to the point of being completely infuriating.
I know it's only on a single benchmark, but I dont understand how it can be so bad...
HN is the last bastion of serious inquiry these days. But its not immune as OPs comment proves.
I really like this benchmarking. Have you evaluated the judge benchmark somehow? I'd love to setup my own similar benchmark.
I haven't evaluated the judge benchmark. You have everything needed in the repo to do so though, so be my guest. It took me a bit of time to put all this together and won't have much more time to dedicate to it before a couple of weeks.
BTW, if you explore the repo, sorry for all the French files...
Your prompt is extremely slim yet you score it on a bunch of features.
The eval prompt is quite extensive: https://github.com/guilamu/llms-wordpress-plugin-benchmark/b...
I personally develop with very detailed spec, and I don’t want nothing more and nothing less compared to the spec.
I found 5.4/5.5 much better at following spec while Opus makes some things up, which aligns with your benchmark but that makes 5.4/5.5 better for me while worse for you.
Input: $5/M tokens at <=272K, $10/M tokens above 272K.
Output: $30/M tokens at <=272K, $45/M tokens above 272K.
Cache read: $0.50/M tokens at <=272K, $1/M tokens above 272K.
Significantly more expensive than Opus 4.7 beyond 272K and at least in my tasks, I haven't seen the model that much more token efficient, certainly not to such a degree that it'd compensate this difference. GPT-5.4 had a solid context window at 400k with reliable compaction, both appear somewhat regressed, though still to early to truly say whether compaction is less reliable. Also, I have found frontend output to still skew towards that one very distinct, easily noticeable, card laden, bluesy hue overindulged template that made me skeptical of Horizon Alpha/Beta pre GPT-5s release. Ended up doing amazing at the time for task adherence, which made it very useful for me outside that one major deficit. The fact that GPT-5.5 is still so restricted in that area is weird considering it's supposed to be an entirely new foundation.
Live decision and heavier agentic evals will continue being uploaded for 24 hours but I don't expect its leaderboard position to change at this point.
GPT 5.5 is the most intelligent public model. And significantly faster than its predecessor.
>API deployments require different safeguards and we are working closely with partners and customers on the safety and security requirements for serving it at scale.
And now this. I guess one day counts as "very soon." But I wonder what that meant for these safeguards and security requirements.
> In 2023, the company was preparing to release its GPT-4 Turbo model. As Sutskever details in the memos, Altman apparently told Murati that the model didn’t need safety approval, citing the company’s general counsel, Jason Kwon. But when she asked Kwon, over Slack, he replied, “ugh . . . confused where sam got that impression.”
Lots of cases where Altman hass not been entirely forthcoming about how important (or not) safety is for OpenAI. https://www.newyorker.com/magazine/2026/04/13/sam-altman-may... (https://archive.is/a2vqW)
There is too much and there are too many, and some of their takes don’t fly if you use Claude daily.
If this sounds crazy to you, though, I'd like to know, and understand why. I miss ChatGPT/Codex.
Cheaper and slower than Opus.
Knowledge cutoff: 2024-06
Current date: 2026-04-24
You are an AI assistant accessed via an API.Just ask it about an event that happened shortly before Dec 1, 2025. Sporting event, preferably.
could be they do it intentionally to encourage more tool calls/searches or for tuning reasons
Knowledge cutoff: 2024-06
Current date: 2026-04-24
You are an AI assistant accessed via an API.
# Desired oververbosity for the final answer (not analysis): 5
An oververbosity of 1 means the model should respond using only the minimal content necessary to satisfy the request, using
concise phrasing and avoiding extra detail or explanation."
An oververbosity of 10 means the model should provide maximally detailed, thorough responses with context, explanations, and
possibly multiple examples."
The desired oververbosity should be treated only as a *default*. Defer to any user or developer requirements regarding
response length, if present.The proper way to figure out the real cutoff date is to ask the model about things that did not exist or did not happen before the date in question.
A few quick tests suggest 5.5's general knowledge cutoff is still around early 2025.
Whatever it is, the cutoff date reporting discrepancy isn't new. Back when Musk was making headlines about buying/not buying Twitter, I was able to find recent-ish related news that was published well after the bot's stated cutoff date.
ChatGPT was not yet browsing/searching/using the web at that point. That tool didn't come for another year or so.
Easiest Turing test ever...
A better test is something like "what is the latest version of NumPy?"
You're probably better off asking something like "what are the most notable changes in version X of NumPy?" and repeating until you find the version at which it says "I don't know" or hallucinates.
Donald Trump won the 2024 U.S. presidential election.They all roughly produce junior developer-level code, continue to have mental breakdowns in their “thinking” stage, occasionally hallucinate things, delete pieces of code/docs they don’t understand or don’t like, use 1.5 times the necessary words to explain things when generating docs and so on.
I'm now testing "avoid sycophancy, keep details short and focus on the facts" in my AGENTS.md files.
AI is like having the greatest developer who ever lived, but she is always on 4 beers.
All the AI players definitely seem to be trying to claw more money out of their users at the moment.
(Note, that stops being true at higher reasoning levels, where our observed total cost goes up ~2-3x.)
30/180 usd on Openrouter. Did I miss something?
4.6 did very well. 90% perfect on first try, got to 100% with just a few followups. 4.7 failed horribly. First produced garbage output and claimed it was done, admitted it did that when called out, proceeded to work at it a lot longer and then IT GAVE UP. GPT 5.5 codex was shockingly good. Achieved 90% perfect on first try in about a fourth of the time. Got to 100% faster and with fewer follow-ups.
I’m impressed.
Would be interesting if you ran your same test with Deepseek v4 and some of the other Chinese models.
I also tried GLM 5.1, it's first attempt was such a disaster I didn't bother working with it any further. It also took by far the longest and wasted a bunch of time/tokens trying to find other converters online (and failing) instead of just reverse engineering the format from the sample files given.
$ uvx swival --provider chatgpt --model gpt-5.5
APIError: ChatgptException
Gave it two very long-running problems I haven't had the courage to work on in the last 2.5 years, solved each within an hour.
- An incremental streaming JSON decoder that can optionally take a list of keys to stop decoding after. 1800 LOC about 30 minutes later, and now my local-first apps first sync time is 0.8s instead of 75s when there's 1.5 GB of data locally.
- Flutter Web can compile to WASM and then render via Skia WASM. I've been getting odd crashes during rapid animation for months. In an hour, it got Skia WASM checked out, building locally, a Flutter test script, and root caused the issue to text shadows and font glyphs (technically, not solved yet, I want to get to the point we have Skia / Flutter patch(es))
If you told me a week ago that an LLM could do either of these, without heavy guidance, I'd be stunned. And I regularly push them to limits, ex. one of Opus' last projects was a tolerant JSON decoder, and it ended up being 8% faster than the one built-in to Dart/Flutter, which has plenty of love and attention. (we're cheating a little, that's why it's faster. TL;DR: LLMs will emit control characters in JSON and that's fine for me, treating them as fine means file edit error rates go from ~2% to 0%)
I just wish it was cheaper, but, don't we all...
Either Opus 4.7 miscounts reasoning tokens, or it's A LOT more efficient than GPT 5.5
I thought they made GPT 5.5 more token efficient than 5.4, but it uses 2x the reasoning tokens.
[0]: https://aibenchy.com/compare/openai-gpt-5-5-medium/openai-gp...
In my place for example, a lot of doctors are using ChatGPT both to search diagnosis and communicate with non-English speaking patients.
Even yourself, when you want to learn about one disease, about some real-world threats, some statistics, self-defense techniques, etc.
Otherwise it's like blocking Wikipedia for the reason that using that knowledge you can do harmful stuff or read things that may change your mind.
Freedom to read about things is good.
I think that's the problem. Who's going to claim responsibility when ChatGPT hallucinates or mistranslates a patient's diagnosis and they die? For OpenAI, this would at best be a PR nightmare, so that's why they have safeguards.
A doctor who gets refused by ChatGPT doesn't stop needing to communicate with the patient; they fall back to a worse option (Google Translate, a family member interpreting, guessing). Refusal isn't safety, it's liability-shifting dressed up as safety.
If there's no doctor, no interpreter, no pharmacist, just a person with a sick kid and a phone, then "refuse and redirect to a professional" is advice from a world that doesn't exist for them. The refusal doesn't send them to a better option; there is no better option, it's a large majority of people on this planet.
Hell is paved of good intentions, but open-education and unlimited access to knowledge is very good.
It doesn't change the human nature of some people, bad people stay bad, good people stay good.
About PR, they're optimizing for not being the named defendant in a lawsuit or the subject of a bad news cycle, it's self-interest wearing benevolence as a costume.
This is because harms from answering are punishable (bad PR, unhappy advertisers, unhappy investors, unhappy politicians / dictators, unhappy lobbies, unhappy army, etc); but harms from refusing are invisible and unpunished.
I think AI proves the contrary. There are plenty of examples of things that are getting worse because of technological advancement, particularly AI. Software quality, writing, online discourse, misinformation have all suffered over the last few years. I truly believe the internet is a worse place than it was 5 years ago, and I can't imagine bringing that to medicine would work out differently.
The medical system shouldn't rely on falling back to crappy workarounds, it should aspire to build the best system it reasonably can.
I had a choice better a doctor that used AI or not, I would much prefer one that did...