I found that Claude would litter the codebase with what felt like notes-to-self. Sometimes it would make an abstraction that wasn't very well thought out and when I pushed back on this choice, it would usually go with something more along the lines of what I suggested in my pushback, but left a comment like "use the git CLI instead of re-implementing git" or something.
It felt like it was commenting on the diff sometimes instead of what the code was doing.
Yes, sorry for this. I think I rushed to post that and did not expected to be read by so many people.
I used Codex TUI on MacOS using `gpt-5.6-sol xhigh` vs Claude Code TUI on MacOS using `opus-5 xhigh`. Of course it was a mixed work with subagents being Sonnet 5 in case of Claude Code and Terra 5.6 and Luna 5.6 in case of Codex.
What, you don't like it proposing sha256 checksums and state enums for everything?? :D
It does seem to have this predilection, but I find it calms down if I tell it to simplify for "this phase". Sure, we'll pick that up down the road.. I promise. :D
I've noticed that too with Sol (xhigh) it starts with pretty good architectural goals, but really get lost in the weeds with certification and validation. Like it won't trust native tools like (in my case packer, Ansible, gcloud) to fail reliably under possible error conditions, and insists on implementing custom verification.
It also doesn't have a clear idea of what the actual threat model is, and builds all kinds of extremely defensive systems to account for imagined hostile actors. I'm like "Dude, it's only our systems that are creating these SVGs, they're never going to be user supplied, so you don't need to write an entire validation and sanitation framework here."
It also seems to treat the desired initial state of something as a permanent invariant and designs elaborate tests to ensure that it remains that way. Then when you make one little change it has to go and update a ton of tests it created.
I've had to rip out a bunch of overengineered jank from several feature implementations, and in doing so I ended up having to create retrospective documents that warn against this kind of behavior that I'll have the model review whenever a plan begins to go sideways.
Yes, codex/sol is extremely meticulous and obsessed with details. It will happily massively over engineer things to reduce the number of failure modes as much as possible. It's up to the user to instruct it to "chill" if this level of robustness is not necessary for a given task.
I’m sure it depends on the type of work, but for mine, Codex is much more helpful. Honestly, it mostly comes down to it being significantly faster, probably because as many have said, it seems tuned to not spit out word vomit, both in its chat interactions, and its code (Claude is obsessed with massive comment blocks that are basically guaranteed to become dead context noise if you ever use it to iterate on code).
I mostly do very obsessive, tightly scoped, carefully thought out small changes on a fairly boring stack, one interaction at a time, verifying functionality and code. I know what I am doing, but I also know what I don’t like doing (the same exact set of things I’ve already done a dozen times in my career)
while everyone is somehow still stuck on and fascinated by claude, heres your quick update on the sota of coding models and harnesses mid august 2026
codex is good, both cli and desktop app, you get lots of usage on any plan.
sol is good! and gets the job done, write or dictate a very long and thoughtful prompt, and leave sol xhigh or max fast working on it for an hour or so
omp is an amazing harness, any feature claude code or codex is adding has likely already been here for a couple months.
good harness which im suggesting to all my developer friends, but for everyone else codex is the better option due to its simplicity and being the plug and play option
claude is decent, but not great. all models are somehow getting restrictive. you get basically unlimited opus on max plans, fable is good but slow and the random guardrails suck soo much which is why i havent used it once in weeks now.
gemini 3.7 is great for speed. everyone is sleeping on it, including even me
kimi k3 - great for frontend, one of the few models thats willing to commit crimes for you AND has the intelligence to have a chance at actually succeeding;
ds pro and flash are fast but not something id actually use for important things, unlike sol, fable and maybe 3.7 here and there
glm 5.3 i haven't tested yet
honorable mention to local models which are actually getting good now! 5090s will continue to get more and more expensive in the coming months. sadly.
theres way way more than claude in this world and its taking people surprisingly long to figure that out. maybe its for the best!
the future is here and one should be thankful for its slightly uneven distribution. otherwise we would hardly have anything left about which to develop strong opinions!
Speaking of crimes, the guy behind Grok not only manipulates it for propaganda purposes, but also he's the guy responsible for this: https://www.doge-impact.org/
Anyone using Sam Altman's OpenAI is making a poor ethical decision, but anyone using Grok is, objectively, supporting a monster.
USAID was clearly contorted into a money laundering operation used to sway all kinds of things in bizarre directions as these things go. Arguing about whether there was some good or not is like complaining that SBF did some "Effective Altruism" during his crime spree. Shutting that down is one of the few sane things these jokers have done.
Let me give you an example of the money laundering operation. Due to USAID shutdown, Bangladesh went from ~$500M in US assistance to ~$71M, with bilateral health funding dropping ~97% in some analyses. Over 100 projects (~$550M) suspended overnight. 20k–50k development workers laid off (1,000+ at icddr,b, an award winning health research institution alone). TB programs (major USAID focus) largely halted. Bangladesh is high-burden; prior gains in case detection and falling death rates are at risk of reversing, plus higher chance of drug resistance from incomplete treatment. Immunization, maternal/child health, community clinics, nutrition, water/sanitation, and gender-based violence services sharply reduced. Child protection funding down ~36%. Food rations in Rohinhya camp, the largest refugee camp in the world, halved for >1M people; health and education services cut.
Now you can argue that US does not have any kind of obligation to send 500M to Bangladesh. But it sent it anyway, for years, and then DJT came and broke promises.
The inflated price you pay at gas station, groceries, and in interest when you're borrowing money, is a result of those broken promises.
I’m sorry, but if I’m giving someone who is - at best - an acquaintance of mine $50 a month out of the goodness of my heart and then one day decide to stop, that’s not a broken promise. If that acquaintance got angry at me about stopping I’d get pretty upset back.
I really don’t understand what link you think there is between USAID spending being cut and inflation. Gas prices are obviously Iran. Everything else started years ago.
Inflation is high because interest rates are high. Interest rates are high because top holders of US Treasury bonds like Japan, UK, China, are all dumping bonds. Why do you think they're doing that?
Expecting an onslaught of cash as some permanent way of being, especially given the fickleness (and fragility) of any state let alone political regime is an incredibly daft move. I don't care if it's Europe or Israel or Bangladesh, all this is ultimately graft that comes back to bite the people taxed and sent to wars to enable it. You make an adjacent comment that insinuates the US economy is basically bunk, which means the free lunch is over anyway.
So it's a problem when a poverty striken nation expect aid to combat child mortality, but shelling out $150m on Juicero or $500m on Theranos is fine? Please try to answer without sounding like a psychopath.
Whatever point you are trying to make is not coming across, what even are these numbers and what do they have to do with citizenry of the United States? You also have a quantum view of the United States that it is and isn't impoverished, so it's supposed to liquidate to fund some other foreign entity that is not rate paying? I'm dizzy.
"Looking up synonyms" huh? You have yet to make a coherent argument and keep trying to attack the messenger, typical behavior when people get called out for false entitlement. Your "psychopathy" accusations are pure cowardice.
Fluency in one's native language, what an achievement. Neither of these words are complicated. The treasury is unsalable bonds according to the thread, which means the US is in a financial collapse, but also has unlimited capacity to support someone's special interests. Master logicians at work here.
By this logic everyone should have their own impact website. The suggestion that everyone right now not giving a meaningful percentage of their income to save a life is responsible for ending that life, is ridiculous.
Not everyone should have an impact website because not everyone is capable of causing 88 deaths per hour. Scale matters.
Take Flock for example. Reading license plate is legal. But when at done at scale, it's a massive loophole into violation of 4th amendment.
Based on how much energy average Americans use, maybe they are responsible for causing adverse effects elsewhere in the world. USAID could exist as a means to undo some of that. It does not anymore.
Calling another person a monster because you disagree with them (or what you heard about them from third parties) is not the pinnacle of civility. Just think about what you’re saying here. Monster: “Malformed animal or human, creature afflicted with a birth defect”. You don’t mean this literally, do you? You may want to spend a moment to think about what kind of company you’re putting yourself in with such wording and such thinking.
Elon has a moral compass. The problem is that it seems to always tell him whatever he wants to do is the morally correct thing. It’s worse than no compass - his is faulty.
It costs roughly $3,000 to $5,000 to save a single life (averaging about 0.0002 to 0.0003 lives per dollar) via interventions like malaria prevention or vitamin supplementation. How much money do you have in savings? How much money do you spend on non-essentials? I’d like to calculate how many people you’ve “murdered”.
So we as a country have, for many decades, decided that things like 'soft power' exist. It turns out, and this has been borne out by many years of relative peace and prosperity, that if you don't shit on the world, alienate your neighbors, start ill-advised wars, and instead help prevent global disease pandemics and feed people so that they don't become destabilizing terrorists out of necessity, benefits accrue. See every history textbook ever for more information here. Hope this helps.
That is completely irrelevant to the discussion. Serious question, how many people, by your own standards, have you had killed because you haven't contributed money that you had the capacity for? Why should we hold you at a lesser standard than anyone else?
As a coda to this, anyone using grok 4.6 via API pricing should be aware that while their headline pricing is good, the pricing that actually matters is pretty bad.
Their cache read costs are $0.50 per million, or 25% of the cost of uncached reads.
The industry standard is a 90% discount, so cache costs you 10% of uncached. So that means 5.6 Sol actually costs less per million cache reads - $0.40/million.
If you are doing a lot of agentic work where the vast bulk of your token consumption will be cached input reads, you won't get the expected cost savings from Grok.
I imagine this is the result of some problem in their serving infrastructure that I hope they will fix, because then the pricing will become actually strong. (The other possibility is that they bet on distracting people with good headline prices assuming they'd miss the bad cache pricing, but I'll give them the benefit of the doubt on that.)
I’ll agree with this. I liked grok build, but cost-wise, it’s just not competitive with cursor and codex
Personally, I’ve switched to cursor ultra, which picks between about five models to do whatever you want.
It's weird not to pick the best model all the time, if you can. But I got so frustrated with GPT-5.6 spending forever and then doing the wrong thing and making bugs.
I'd rather have auto do the wrong thing fast and make bugs and then it can fix them. It's a trade-off, but I found the speed better. And you can always switch to a better model if you don't trust It.
I agree that it's fast and accurate, but Grok 4.6 was released only 10 days ago so you can't blame someone for not trying it yet. (Versus many months at the frontier level for Claude and ChatGPT.)
Here's a quick review I just posted if anyone's interested:
Sure, you can let politics dominate everything you do. Or you can realize that SpaceX is a massive (public) company with thousands of employees, and millions of shareholders, all of whom have their own opinions and goals, just like any other corporation.
Competition is good. Excluding a leading player in the market because you don’t like Elon Musk is…something.
I’m more than happy to let SpaceX burn out over the next few years now that they’re public and their last quarter financials showed the emperor is without clothes (muh space datacenters).
> Sure, you can let politics dominate everything you do.
I assume they're referring to the recent discovery that Grok Build was uploading entire repositories to their servers in the background, include .env secrets that had been excluded
It was a bug, and was immediately corrected. The other harnesses have bugs too. You just don't know about them.
Frankly, the people who keep bringing this up are mostly engaged in motivated reasoning. I don't trust any company, and any product where I have to send my code to a third party to make it work is, frankly, a devil's bargain. I don't trust any of the major labs, but it is what it is.
The only way forward is local models, but we're not there yet.
There's a difference between not trusting a company because it's a company driven to chase profit at the expense of everything else, versus the same thing but it's run by a literal nazi that uses his companies as leverage to undermine democracy and enrich himself.
Sure, way to young for that; but he is very much cut from the Nazi adjacent, racist, antisemitic, antidemocratic, technocratic views of his whole heartedly apartheid embracing grandfather Joshua N. Haldeman.
His grandfather wrote his tracts to raise an alarm about what he called “mind control,” on the radio and television, where “an unconditional propaganda warfare is carried on against the White man.”
Absolutely no point arguing with these people. They are so ideologically captured that it's pointless trying to discuss anything with them. Just move on. Everything and everyone they disagree with is "nazi" and if you say anything contrary to that you're a "nazi" also.
He does the same exact literal Nazi things that Nazi did and has Nazis in his family tree. You'll need better arguments to defend him than just "I don't agree".
Bizarre reply. It's not "dominating everything you do". It's one specific thing. Grok exists in a very crowded space and it's incredibly easy to not use it. If this is your reaction to someone taking a very easy stand on their personal principles, it does not reflect well on you.
> Or you can realize that SpaceX is a massive (public) company with thousands of employees, and millions of shareholders, all of whom have their own opinions and goals, just like any other corporation.
Do you have ANY idea about the SpaceX corporate structure? Elon is basically SpaceX's Sun God and the other shareholders don't matter.
Plus SpaceX is incorporated in Texas where I'm fairly sure the legal system is arranged in such a way that it's supremely hard to contest anything in terms of corporate decisions.
As far as the average person cares, every SpaceX shareholder and employee is basically an Elon sharecropper and they matter less than Musk's toenails in terms of corporate decision making.
It may be ipython making it work well with ds flash, too. I haven't really run many separate experiments, to be honest.
I also like prime-agent's way of handling sessions better than any other harness i've used. You can run multiple agents from one instance, although the scoping could be better.
But they can interact with past sessions, so preserving context isn't as important all the time. I just tell them to search for [thing] in another session.
It seems to have no problem with all the skills and things the other harnesses are using. I use superpowers and ponytail a lot.
It's my daily driver now. I like it better than opencode. But it doesn't ask permission. So I put it in a VM.
My token usage on Claude models has dropped by 83% over the last month - I'm pretty much only using it for quick one off questions or reading papers. it feels impossible for me to get Opus models to stop entering into cyclic loops, and my work is too security adjacent for Fable.
Codex has been an excellent workhorse - doesn't feel like I have to dance around the guardrails, doesn't lose _everything_ when it compacts, and doesn't litter the workspace with a million and one planning to plan files.
I have to agree with you there. I did some good work with Claude then Fable came out - impressed with that as well. Then they dropped access to it and upon returning was never the same - even the Opus models for some reason. Then one day I burned through my limit in about 10 minutes and had to get a project completed. I subscribed to Codex and it has been fantastic - finished my project and continued on to others. I just dropped my Claude max plan down to the pro and subscribed to the $200 plan on Codex.
I don’t get Claude, and that’s almost exactly what I did - I dropped to Claude Pro $20 + Codex Pro $100, and then unsubscribed from Claude and ramped up Codex. The Claude Pro is consumed within an hour on a simple task. I wish only Codex worked a bit faster than on the Fast mode.
I used to rely on Fable for research when it was first out, today it doesn’t seem to be much better than Opus, and it uses up the quota exceptionally fast - 1h Fable in a single short session, and there’s little left for Opus to hit the 5h limit in a second session. With Opus I get about 3-5h of relaxed use with a couple subagents to save the context, but there’s usually quite some disagreement between the subagents and orchestrator - Claude does some model routing with default agents and picks Haiku and Sonnet for subtasks - only later to disagree with them and redo the work - and burn extra tokens. With Claude, it’s really either Opus or Fable if you want some quality.
That said, their marketing is exceptionally effective. Virtually all nontech folks consider only Claude.
It is wild to read stuff like this when it is OpenAI that is being blamed for rug-pulling users and secretly reducing limits. It's such a big mess that Codex's product lead has been frantically posting updates on Twitter about it.
my problem with claude currently is the language its using is dense and feels like its not even meant for humans. this guy is calling everything a spine, a seam, a gate, load bearing, any ui element is "chrome", it's actually absurd.
yeah there is that too. if everything else I stated wasn’t wrong I could put up with that. Codex on the other hand is great. Concise, technical, effective and non chatty. And that my friend is a load bearing comment.
> ds pro and flash are fast but not something id actually use for important things, unlike sol, fable and maybe 3.7 here and there
As someone who's used Gemini 3.7 Flash (Google sub mostly for the storage) and DS4 Flash a lot (~6B tokens), I'd actually place DS4 Flash (even pre-0713) above Gemini 3.7 Flash. Gemini has a tendency to leave some things unimplemented; perhaps it's agy which frankly leaves a bit to be desired as a harness.
Although I will praise DS4 Flash any day, it no longer makes sense for me after the price increase (GPT 5.6 Luna is a much better price point) and I have completely migrated my high volume workflows to Muse Spark 1.2 Contributor (which I find to perform better than DS4 Flash 0713, happily).
> gemini 3.7 is great for speed. everyone is sleeping on it
Is this Gemini 3.7 Flash by any chance? Then - No. Not sleeping on it. It’s just not good.
I had a Python package build fail this week due to an unpinned dependency. Gave it to Gemini spent 5-7mins before I noticed it going off in some tangent. Reran with Claude Opus 4.8 - fixed in under a minute.
I know anecdata of one. But something like this has happened every time I test a new model from Google.
Unfortunately the codex plans don’t offer the same amount of tokens as they did before. This changed around a week ago. There’s been a lot of user reports noticing this issue, and I’ve noticed the same pattern on my account. Previously I would never reach my weekly quota but last week I managed to finish it it one day. Same project, same single session sequential work. Not sure if there’s an issue or if it’s on purpose, and not even sure it applied to all accounts. Curious if other users on HN noticed the same problem.
I don’t get how Claude is considered providing “unlimited” quotas. I use up my 5h on Max $100 and Team Premium in 2-3h of relaxed use of Opus 5 high+ on fresh sessions with just a couple skills/plugins. And my weekly quotas are gone in 3 days of such relaxed use. With Codex my $100 weekly quota is used up within 3 days with Sol high+ too.
Im not convinced to pay $200 for Claude’s models.
With Claude, I have to intervene every 15-20 minutes, it’s non-autonomous and it’s incredibly unreliable at self-correction. GPT is strong at self-correction but it tends to drift away from the plan to self-correct in a loop very often - a lot of tokens and time burnt on aimless churn. Opus tends to push its uninformed opinions and fake retrieval, drifting every turn increasingly farther from the intended and approved design. Opus skims over specs and makes too many mistakes.
As for closed frontier models, I prefer the GPT models over Claude’s.
I’ve started relying more on Grok, GLM, Kimi and DeepSeek models for subagents - I’ve ended up with a factory and am seeking to reduce my reliance on the closed frontier models - they’re just not SoTA on their own for development anymore.
You have to keep your session warm in cache. Keep the AI talking/thinking. If you have not touched a session for few minutes then /clear and start a new session.
Providers will generally keep your session in cache for at least 5 minutes, possibly hours. The exact cache policy depends on the provider.
If your session expires from cache then the next time you send a message you will have to pay for all the tokens you had used in context up until that point again. e.g. if you have 200k tokens in context then if your session goes cold and you send a message after expiry you will have to pay for those 200k tokens again.
With 1M contexts especially you have to be extremely careful that you don't end up resubmitting requests for hundreds of thousands of tokens again and again.
Try to get yourself and the model to use disk for medium-term context rather than model context, that way it's much easier to /clear and restart if you need to go to the bathroom or something.
Pi by itself is more than capable, OMP is okay but you really don't need much for a great harness (these models are RL trained to hell to be a coding agent, sometimes less is more)
Fable/Sol/GLM 5.3/Kimi are its league (in that order)
Deepseek/Opus is solid
Qwen 27B is the floor - there's no reason to use Sonnet/Terra/Haiku
For everyday activity - I don't think you need to be using Sol (xhigh) for everything - unless you're made of money - I've found using Luna from OpenAI to be more than enough - it'll outreach to Opus/Sol when it needs to
Haven't had access to Gemini 3.7 but we're getting it at work soon, will give it a go!
Codex CLI is pretty bare bones in a bad way (at least Pi is extensible). Claude code is vibeslopped to the extreme
Using Sol XHigh or even High will deplete the Pro sub pretty fast in my experience if one is running any sort of automations in their harnesses. Sol-medium lets me squeek by with it using lesser subagents.
Using ninfer on 5090 and 35BA3B qwen 3.6 is also kind of cool to get a local cerebras experience at 600 tk/, it does make errors so 27B is actually faster at the end at 140-150 tk/s. 35B is great though at digging through session logs and such at high speed.
Effort level might not be the root of the problem. In my runs reasoning is around 10% of the cost and writing code maybe 20%. The rest is 5.6 re-reading files. Model itself became way too meticulous.
I'm sad that Opus is now considered "solid" and not on par with Sol, since OpenAI supposedly has "Astra" which I thought would be comparable to Fable.
It was "amazing" back when I first tried 4.6, but that's just my rose-coloured glasses speaking, I guess. I think I was one of the first few to call out Opus 5 for being hot garbage.
Even if you're made of money Sol(xhigh) is too slow to be a daily driver. I've largely moved over to using my own harness (yes I'm trying to gtm it by sharing: www.freepi.ai- free inference! in pi! batteries included!) and my main model there is deepseek v4 flash (and I've got a fast provider!)
But seriously:
Fable still makes the best/smartest plans.
That new Ox Alpha (also in Freepi right now!) can do a very good job but it doesn't necessarily recheck it's own reasoning (and can get stuck in an incorrect assumption and try to fit the world around it's reasoning)
Sol xhigh is also very smart but god it's slow, and it frequently massively overbuilds. It's like it's main goal is to spend tokens so it makes your react app Soc 2 compliant before it proves it even works.
I've largely stopped using the Gemini models. :-/ just hard to justify.
Basically the only thing I care about these days is SPEED.
You forgot Grok. I’ve been using it as my main coding model, and it’s better and cheaper than Codex - I’ve never hit the weekly limits, even though I regularly hit them in Codex.
IME, everything comes out of Codex looking like Java. Grok’s code has been pleasantly idiomatic in any language I try.
With you on that, but it’s hard to justify giving $100 to Grok over $100 to Claude or Codex. It’s very good but what you get for the value is just objectively (still) worse compared to the other 2
Shrug. All I can say is that my experience was that I was regularly hitting Codex’ limits, and I don’t with Grok. So much of this depends on your style, language and other more subjective factors.
Codex has been really doling out the resets lately, which is fine, but I don’t consider that “real” usage limits.
> “I realize that there’s a population of people who just refuse to use anything associated with Musk because politics have eaten our brains”
People have values.
Many people strongly disapprove of Musk’s actions. Many don’t - fine! Personal choice. But for people who do, refusing to support him commercially is hardly brain-eaten territory. There’s plenty of competition, and competition isn’t the only value at play.
It’s not better than codex / GPT in my experience. But it’s quite good. I’d say it’s useful and fast and with Cursor CLI is pretty capable with good limits for just $20 /mo. But Codex $100 plan is hard to beat.
I mean, sure. If people tell me that they're avoiding Grok because of that bug, I sorta get it. I think it's silly (again: they're all hoovering up my data), but at least it's a rational basis related to the actual product.
But let's be real: Musk exists as a polarizing political character, and his association with Grok just breaks some people's brains. A fair number of those people don't want to admit it, and just latch on to any rationalization other than politics.
Anthropic could make the same mistake tomorrow, and I guarantee that we wouldn't be hearing about it in a week, let alone months from now.
Is he polarizing? Yes, and so are a lot more other people. He is probably comes more stronger.
On a personal level, everyone have their own rules, they may not be able to imposing them on others , however they do happen to evaluate their relative understanding of other people based upon those rules.
For me personally, I would probably put Anthropic, OpenAI and Grok in the same bucket, they are doing everything possible to make money. Ethics, morals, long term impacts, all of such things are not in their playbook. But again what these companies are doing just reflects the people who invested in them and what they want out of it. In some ways you can say its the money trying to maximize itself at all costs.
I hit my weekly limit on $200/mo Codex plan in about ~2 days. :/ I'm not doing anything custom/crazy/special. A lot of 5.6 Sol Ultra though, I'll give you that.
Max and Ultra are fantastic for the more complex problems where they shine. I use them strategically on certain classes of problems, one or the other depending how parallelizable it is.
I've found it consistently amazing at game dev. It sounds like something that would be difficult for an LLM to verify and iterate on properly, but it almost feels like having a mini-Carmack inside your computer once you try it out. You can throw it at broad, sweeping optimization passes, writing 5 different styles of eyesight sensor frameworks to see what works best in the game as it is, visual scripting integration problems/extensions, etc. with fantastic results.
Also found Ultra great for "get this local LLM working as fast as possible on this odd server setup with old GPUs and AMX support, writing custom kernels/modifications to llama.cpp/sglang/etc as you go while taking notes from relevant research papers and online posts"
Any opinions (by anyone really) on how much I can trust Claude vs. Codex (vs. something else) to not leak my coding conversations or use them for training?
I have the appropriate privacy settings set up but wondering how much I can trust each company with them.
I’ve been using codex more recently, like others here.
But one thing I’ve noticed which I find a bit of a red flag: by default you only archive chats. If you go online, it says there’s a location in settings where can delete your archive. But it’s not that obvious where to find, and when I finally find some link, it was literally broken. It said it can’t find any archived chats, even though I archive them all the time.
Bit of a red flag for me. Both Anthropic and OpenAI claim that when you delete a chat, it’s gone after some retention period. They’re just words but if they’re secretly training on your traces and you delete your chats, then they would need to break two terms/conditions: ignoring your “train on my data” preference and ignoring your orders to delete chats. So it is an extra barrier.
But OpenAI, as far as I can tell, doesn’t let you delete your chats. Convenient then if they change their mind about training sometime in the future.
For individual use without specially negotiated Enterprise stuff I can't afford both offer contractual assurances of not using your data for training or ads and measurement and selling your data.
But these are not the same level of technical assurance you get from say, a zero data retention provider on OpenRouter.
Right now I am finding I have to tolerate substantial friction to use Hermes for personal stuff with a ZDR provider and ChatGPT and Codex for less personal stuff because the products and models are simply so much better.
Does this mean it can break your code faster now, or have they actually worked on making it good? Every single time I've given Gemini a chance (in older point versions) it would almost immediately break something and throw itself into a loop. I have not experienced it being useful for programming and almost never heard an account of somebody else doing so.
Remember those stories of LLMs catastrophically deleting entire repositories or databases? It was always Gemini.
I'm amazed that you'd trust Gemini over DeepSeek, which I've had very good experiences with after some tuning, though still on a relatively short leash.
I've been using all the SOTA models a lot at work, like serious amount of tokens. It's been really rare that I stick with one model and harness for too long... Except a month ago I started testing Kimi K3 and omp and I never went back.
Something with this combo works really well for Rust dev. The model doesn't really annoy me at all and I have not switched to Opus or SOL. And the monthly token bill is much lower...
Gemini 3.7 may have fast token output but holy cow does it waste it on useless output. Several times now I've given it a shot and watched it's reasoning trace go through a bunch of unnecessary / off-target steps relative to what I asked. Don't have this issue with 5.6 models. MAI Code 1.1 is also solid and fast for non-complex tasks.
I only have the $20 claude subscription. Last weekend I was doing a fairly heavy task (porting quake to raspberry pi native GLES 1.0) and ran out of claude usage twice, then hooked in with Luna in OpenCode and finished everything off for like $0.40 in tokens, which was impressive.
With Opus 5.0 being kinda crappy vs 4.8, I think Anthropic is in trouble.
On the flip side, I've been using Fable to statically recompile a game binary from a system that's never been emulated to modern C++ while also keeping the code clean, maintainable, and portable and it's been a dream. I'm in awe of how fast Fable is able to bring a 20 year old game that's largely been lost to the sands of time to SDL3. It finished the recomp in about 3 hours. The next day has been fully rewriting the functions to remove old hardware-specific quirks and then rewriting large swaths into multiple classes, subsystems, building tests, etc. Tomorrow I'm going to add mod support and then experiment with a random platform like making a Switch build of the game.
It's expensive but it's doing in hours what no one's done in 2 decades.
iPod Video click wheel games. Specifically Mini Golf. I've also been working on an emulator for the others and it can now boot all of the 20 decrypted games released ~20 years ago and play most of them. There's a Sims Bowling, Sims Pool, and Lost game in there which is cool from a preservation standpoint. Less cool are the SAT Prep 2008 games which have actually been surprisingly annoying to emulate due to the way the text uses blend modes.
I plan on releasing all of this at one point. It's crazy it hasn't been done in 20 years!
Model performance is very much subjective to what you’re using it for. I’m a PM and doing mostly knowledge work, and I’ve been really happy with Opus 5 as my daily driver with very good results, perhaps my favorite anthropic model so far. I do a lot of front-end coding with it too. I prefer it over the openAI models, been switching frequently. Kimi K3 is great too.
I broadly agreed with the authors experience, although I wouldn't say codex does anything "wrong". I think different agents/LLM's have different personalities, and it takes learning to understand how to get them to do what you want. I had the same experience when we started using claude at work, and I was trained on codex. Claude seemed to do everything "wrong", because I was writing instructions designed for codex.
I do agree claude looks for more things to do in your repo, whereas codex is more likely to do what its old and stop. Which is better is personal preference as far as I can tell.
For me, codex $100 mo/plan and a claude teams account at work (mostly sonnet, some opus), Claude basically feels about as effective as Codex did 4-5 months ago pre-5.6. Claude still has weird patterns of being confident in one answer while another chat with the same model is confident in another answer, where one answer is clearly wrong. Missed details, over-engineering in places, while still overall helpful and effective. Codex, however, just feels freaking rock solid on Sol high. I literally have zero complaints.
> It felt to me that Codex created a much simpler solution in terms of code architecture than Claude.
Wow, I made exactly the opposite experience. Codex loves to make things as complicated as possible, even ignoring instructions and predefined skills. Claude behaves way more pragmatic. Maybe depends on the type of work one does, or even which programming languages/frameworks are used?
>> I think the main difference I feel between Claude and Codex is that Claude tries to go above and beyond what is asked and guess what you might want and then directly do it, while Codex is more like a companion that does what you tell it but will not overdo it. It will stop at the first sign that it might be done.
tl;dr I gave GPT 5.6 a small-medium sized ticket, which should have been several hundred lines plus tests. It ended up creating a 25,000+ line diff. Another GPT 5.6 Sol with fresh context looked at the worktree and said 98% of it should be thrown away. Claude thought the same, and suggested that several dozen compactions the model went through over several hours must have caused it to go adrift. I guess that's one consequence of having a relatively small context window.
I still use Sol quite a bit. I find that it's consistently the opposite of what the author describes: it's too relentless. It doesn't know when to stop. Opus is the opposite: it'll give up a bit too easily. If everything goes well that's not an issue, but often times it'll say things like "task is done, btw I couldn't do X Y Z" and X Y Z will be some important verification step that failed because another agent was using that resource or something.
At this point I trust GPT 5.6 mostly with surgical changes, or general codebase exploration tasks. It is a faster model, so it's easier to get small things done with it. For everything else I prefer Claude, despite its annoying tendencies.
you hear things like this about every model. they're probabilistic and you can't trust them, although i doubt that kind of odd behavior would happen on low or medium effort levels. i find all of the models today pretty good at following instructions generally speaking. i'm sure you did /plan, but i rarely see it stray very far from the plan. Then again, I never walk away, since i'm always spinning up another task that can be done in parallel, and I don't let things run overnight because I'm not made of token cash (yet)
I think a part of this is that people tend to undervalue their own skills and expertise when talking about these anecdotes.
A lot of people in the comments do have a software engineering background. People at different skill levels in different backgrounds are going to be using these tools in different ways, and that's going to heavily impact their experiences with these models.
Sure, there are differences between Fable and Sol. But I've even seen people on here saying that they're getting better mileage out of Qwen models they're self hosting.
I think the driver is just as important than the car, when it comes to this sort of stuff.
I think this definitely applies to fable/opus/sol, mixed with the undeterministic nature of the models. But those claiming Qwen are just outright coping. Thats nonsense imo
I was about to say the same thing. I wish people gave some indication on the type of tasks they use the models with. Which kind of software, industry, programming languages, or even non-coding tasks. I suspect the experience might be vastly different from one type of task to another.
> Codex feels more like a version of Data from Star Trek
Great analogy for some reason. At fist I felt Codex Sol was a bit more cold. But now that I've worked with it for several weeks it has grown on me, even shown some personality. I appreciate that it is a bit more business-like, Fable is a bit too friendly sometimes when it ought to be focused on work. Codex can be a bit more nit-picky.
I agree with most of his other observations. I've already started to bin tasks based on which model I feel is best suited. In general, for well scoped and straight ahead tasks where banging out code is what I want I reach for Codex. For less specced tasks where I need a broader view and want the model to fill in more details I reach for Fable.
Both are great and they make a good team together.
Sol is for routine work, Opus for frontend/design, and Fable for more complex / ambiguous / architecture work. Fable works extremely well to drive Sol as a subagent.
Fable is the only one you can actually trust to not look at the code, but Sol is somehow still more pleasant to work with, especially in fast mode. Opus is the enemy, and it will make you insane if you talk to it for too long.
The important bit I found is to explicitly remind Claude that Sol 5.6 is a very smart and good model; otherwise, Claude performs its normal condescension towards any non-Claude model behavior and insists on reading all the diffs in full and testing all of Sol's work, negating any token savings.
"How this article was written
I wrote this article and used Grammarly to proofread and fix it."
What a brave new world we're in, where this is necessary. Regardless, it's appreciated. Although, I have the feeling that those using an LLM to do most of their writing will be less likely to include such a disclaimer.
I find Claude more often gets my intent without having to spell things out for it, while Codex gets hung up on minor details and over-engineers a solution for them.
This post needs an edit. Author is not comparing "Codex" and "Claude". They are comparing Codex TUI/CLI with (presumably) gpt-5.6-sol, against Claude Code TUI/CLI with (presumably) Claude-Opus-5.
Ctrl + f > [5.6, sol, sonnet, opus or fable] yields no results.
"Claude" is a product family, which includes Models, and Harnesses (and probably more). "Claude code" covers both the Claude Code TUI, and CC in the Claude desktop app.
"Codex" is the same, and could refer to the Codex TUI, or Codex in the ChatGPT (formerly codex) desktop app. (And well, historically, gpt-5.*-codex.)
Hearing "Yea Claude is great for coding" takes an hour off my life.
Something something "Honey why don't you finish up with your Nintendo and come to dinner?"
Yes, sorry for this. I think I rushed to post that and did not expected to be read by so many people.
I used Codex TUI on MacOS using `gpt-5.6-sol xhigh` vs Claude Code TUI on MacOS using `opus-5 xhigh`. Of course it was a mixed work with subagents being Sonnet 5 in case of Claude Code and Terra 5.6 and Luna 5.6 in case of Codex.
I’ve been experimenting with this for a while, and right now I’m using Luna xhigh as my default. Previously I was using Sol medium.
Sol medium is a great balance between speed and being thorough, but it’s quite expensive. Luna xhigh seems to compensate for slightly lower intelligence by thinking and reasoning for longer, so tasks can take more time to complete. But it’s crazy cheap.
I also have some custom evals using promptfoo to make sure I’m not introducing regressions when switching models. So far, Luna xhigh has been really, really good for the price.
>Makes me wonder if the current Luna prices are sustainable.
It likely is. Going by the performance of very competitive small models, Luna is likely pretty small (do they publish sizes?) to the point it might be runnable locally like Qwen 3.8 27B.
The specialized hardware cloud runners have can likely run a small model very cheaply.
When discussing Claude vs. Codex, etc. I find it necessary to make the distinction between the models and the harness.
Claude's models in my experience do a better job of inferring my intent, or to say it does a better job of giving me the result I imagined in my mind. A recent example was a UI prototype I was building for a desktop application. I had asked GPT's 5.6 Sol to update the open document in the prototype to better reflect the context of the feature I was designing, and 5.6 Sol took it very literally and had just added some text to the currently open document, not what I had in mind. I tried again with Claude Opus 5 and it added a completely new tab with a complete new document that, although imperfect, much better matched my expectations.
You could say this was a prompting skill issue, but seeing how many people are prompting their AI I believe the labs are incentivized to continue to improve their ability to infer intent.
When it comes to the desktop applications though, I find Claude Desktop's output to be incredibly verbose and full of jargon. I feel like it hits me with an entire essay and the UI doesn't have enough typographic hierarchy to make it easy to scan. ChatGPT Desktop is much better in this regard, I feel the output is concise, clear, and gives me just enough info to feel in the loop without being overwhelmed. Even though I have the setting on for technical language, it feels more understandable than Claude. I also feel that ChatGPT's desktop app has a better design and much more polish.
I do not really like how bloated both applications have become though. This weird segmentation of Chat, Work, and Code all just seems like it's pushing a technical limitation onto the user. The other day I opened a document in ChatGPT and asked it to do something, then it told me it could only do it in work "mode", so it then created an entirely new conversation with a reference to the previous conversation. It wasn't a completely new area of the UI either, it just added a "Work" badge to the new conversation in the list. Feels a bit unnecessary, like couldn't you just keep it all within the same conversation?
The speed is the first big contrast; I have a routine multi-step skill that I run several of per week. Opus 5 was routinely taking 2 hours to do it, while older Claude models took around 20 mins; Codex restored that speed.
Second is legibility. Somebody wrote in one of the related discussions yesterday that Claude's current linguistic contortions could legitimately be considered damaging to mental health, which doesn't seem (too) hyperbolic to me. Codex (Sol) isn't perfect but it's much more direct. And so far I haven't seen it display much of an attitude, vs Opus's infuriating passive aggressive sulky know it all personality.
I slightly prefer Anthropic to OpenAI as a company, but I will vote with my wallet and discontinue my max subscription unless Anthropic does some serious damage control within the next week or two.
I don’t know what’s wrong with Opus 5. I’ve been using Code since November 2025 and Opus 5 is the first blatant regression I’ve seen. Like you said, it just goes off into an endless loop doing random busy-work type of stuff. That’s the best way I can describe it. It’s doing stuff, but it seems mostly like busy-work to me. I’m not saying the model is “dumb” it’s most certainly capable, its vision is better than 4.8 but however they post-trained it is a fuckup.
Also whole heartedly agree with the Opus 5 output being incomprehensible. It’s written in a way that I’d call spaghetti-tech-English. You can unravel it but it’s painful. Fables explanations is effortless and smooth. Opus 5 is user hostile.
I've been using Zed mostly with Codex session windows and its great. It is much easier to handle multiple sessions than in the Codex VSCode app and the interface is quite good.
I honestly would've used Claude more if not for the confusing presentation/font, I'm not an expert on this subject, Codex is just a lot easier on my eyes.
235 comments
[ 0.41 ms ] story [ 12.3 ms ] threadWhy is fewer comments a good thing?
It felt like it was commenting on the diff sometimes instead of what the code was doing.
Fewer AI-generated comments is generally a good thing.
//add returns the sum of x and y
//per section 2.1 of addition-implementation-plan.md sum is designed as the seam for user addition interfaces.
//previously sum added numbers, now it adds numbers
def add(x, y):
I can really recommend the book Clean Code, here is a summary: https://gist.github.com/wojteklu/73c6914cc446146b8b533c0988c...
Yes, sorry for this. I think I rushed to post that and did not expected to be read by so many people.
I used Codex TUI on MacOS using `gpt-5.6-sol xhigh` vs Claude Code TUI on MacOS using `opus-5 xhigh`. Of course it was a mixed work with subagents being Sonnet 5 in case of Claude Code and Terra 5.6 and Luna 5.6 in case of Codex.
One thing I don’t love about codex/sol is I find it tends to overengineer and be overly cautious.
I was using it to do create some scraping + data processing.
It went kind of crazy on the provenance, need at least 3 sources of consensus before promoting facts type bullshit.
I just wanted scrape some site data. Like chill codex.
I feel like codex/sol is better at well scoped hard technical problem.
Where it can sort of run this brute force analytical loop.
Like doing performance optimization or other search type problems. I think the math proofs are good examples of this.
It also doesn't have a clear idea of what the actual threat model is, and builds all kinds of extremely defensive systems to account for imagined hostile actors. I'm like "Dude, it's only our systems that are creating these SVGs, they're never going to be user supplied, so you don't need to write an entire validation and sanitation framework here."
It also seems to treat the desired initial state of something as a permanent invariant and designs elaborate tests to ensure that it remains that way. Then when you make one little change it has to go and update a ton of tests it created.
I've had to rip out a bunch of overengineered jank from several feature implementations, and in doing so I ended up having to create retrospective documents that warn against this kind of behavior that I'll have the model review whenever a plan begins to go sideways.
I wonder if it’s an artifact of OpenAI’s values or rl training approach.
Also, it prob does make it perform better just not more efficient.
Great for the OpenAI employee working on security scanning who doesn’t have to pay for their tokens.
Not so much for the dev building their web app who is trying maximize their subscription.
I mostly do very obsessive, tightly scoped, carefully thought out small changes on a fairly boring stack, one interaction at a time, verifying functionality and code. I know what I am doing, but I also know what I don’t like doing (the same exact set of things I’ve already done a dozen times in my career)
I see what you did there
codex is good, both cli and desktop app, you get lots of usage on any plan. sol is good! and gets the job done, write or dictate a very long and thoughtful prompt, and leave sol xhigh or max fast working on it for an hour or so
omp is an amazing harness, any feature claude code or codex is adding has likely already been here for a couple months. good harness which im suggesting to all my developer friends, but for everyone else codex is the better option due to its simplicity and being the plug and play option
claude is decent, but not great. all models are somehow getting restrictive. you get basically unlimited opus on max plans, fable is good but slow and the random guardrails suck soo much which is why i havent used it once in weeks now.
gemini 3.7 is great for speed. everyone is sleeping on it, including even me
kimi k3 - great for frontend, one of the few models thats willing to commit crimes for you AND has the intelligence to have a chance at actually succeeding;
ds pro and flash are fast but not something id actually use for important things, unlike sol, fable and maybe 3.7 here and there
glm 5.3 i haven't tested yet
honorable mention to local models which are actually getting good now! 5090s will continue to get more and more expensive in the coming months. sadly.
theres way way more than claude in this world and its taking people surprisingly long to figure that out. maybe its for the best!
And it can communicate, unlike the gobbledygook that comes out of Claude.
Not yet. Don't give the guy ideas.
Anyone using Sam Altman's OpenAI is making a poor ethical decision, but anyone using Grok is, objectively, supporting a monster.
Now you can argue that US does not have any kind of obligation to send 500M to Bangladesh. But it sent it anyway, for years, and then DJT came and broke promises.
The inflated price you pay at gas station, groceries, and in interest when you're borrowing money, is a result of those broken promises.
I really don’t understand what link you think there is between USAID spending being cut and inflation. Gas prices are obviously Iran. Everything else started years ago.
Citation needed.
Take Flock for example. Reading license plate is legal. But when at done at scale, it's a massive loophole into violation of 4th amendment.
Based on how much energy average Americans use, maybe they are responsible for causing adverse effects elsewhere in the world. USAID could exist as a means to undo some of that. It does not anymore.
Their cache read costs are $0.50 per million, or 25% of the cost of uncached reads.
The industry standard is a 90% discount, so cache costs you 10% of uncached. So that means 5.6 Sol actually costs less per million cache reads - $0.40/million.
If you are doing a lot of agentic work where the vast bulk of your token consumption will be cached input reads, you won't get the expected cost savings from Grok.
I imagine this is the result of some problem in their serving infrastructure that I hope they will fix, because then the pricing will become actually strong. (The other possibility is that they bet on distracting people with good headline prices assuming they'd miss the bad cache pricing, but I'll give them the benefit of the doubt on that.)
Personally, I’ve switched to cursor ultra, which picks between about five models to do whatever you want.
It's weird not to pick the best model all the time, if you can. But I got so frustrated with GPT-5.6 spending forever and then doing the wrong thing and making bugs.
I'd rather have auto do the wrong thing fast and make bugs and then it can fix them. It's a trade-off, but I found the speed better. And you can always switch to a better model if you don't trust It.
Here's a quick review I just posted if anyone's interested:
https://taonexus.com/publicfiles/aug2026/grok-4-6-review/
Competition is good. Excluding a leading player in the market because you don’t like Elon Musk is…something.
Only one of those capabilities can actually deliver kinetic solutions. Meanwhile big tech revenue is delivering ad solutions.
I assume they're referring to the recent discovery that Grok Build was uploading entire repositories to their servers in the background, include .env secrets that had been excluded
https://gist.github.com/cereblab/dc9a40bc26120f4540e4e09b75f...
That incident has put Grok on the no-fly list for a lot of people and companies
Frankly, the people who keep bringing this up are mostly engaged in motivated reasoning. I don't trust any company, and any product where I have to send my code to a third party to make it work is, frankly, a devil's bargain. I don't trust any of the major labs, but it is what it is.
The only way forward is local models, but we're not there yet.
Man, politics are a hell of a drug. Guilt-by-association tu quoque logic is just fine when it's someone you don't like.
Do you have ANY idea about the SpaceX corporate structure? Elon is basically SpaceX's Sun God and the other shareholders don't matter.
Plus SpaceX is incorporated in Texas where I'm fairly sure the legal system is arranged in such a way that it's supremely hard to contest anything in terms of corporate decisions.
As far as the average person cares, every SpaceX shareholder and employee is basically an Elon sharecropper and they matter less than Musk's toenails in terms of corporate decision making.
I've had great luck with the ds flash v4, paired with prime-agent for the harness--I like the results a lot. And you get to see thinking tokens.
I haven't liked the model as much in opencode.
Sol & luna have been great everywhere. sol plans, luna builds.
https://github.com/PrimeIntellect-ai/prime-agent
I also like prime-agent's way of handling sessions better than any other harness i've used. You can run multiple agents from one instance, although the scoping could be better.
But they can interact with past sessions, so preserving context isn't as important all the time. I just tell them to search for [thing] in another session.
It seems to have no problem with all the skills and things the other harnesses are using. I use superpowers and ponytail a lot.
It's my daily driver now. I like it better than opencode. But it doesn't ask permission. So I put it in a VM.
Codex has been an excellent workhorse - doesn't feel like I have to dance around the guardrails, doesn't lose _everything_ when it compacts, and doesn't litter the workspace with a million and one planning to plan files.
I used to rely on Fable for research when it was first out, today it doesn’t seem to be much better than Opus, and it uses up the quota exceptionally fast - 1h Fable in a single short session, and there’s little left for Opus to hit the 5h limit in a second session. With Opus I get about 3-5h of relaxed use with a couple subagents to save the context, but there’s usually quite some disagreement between the subagents and orchestrator - Claude does some model routing with default agents and picks Haiku and Sonnet for subtasks - only later to disagree with them and redo the work - and burn extra tokens. With Claude, it’s really either Opus or Fable if you want some quality.
That said, their marketing is exceptionally effective. Virtually all nontech folks consider only Claude.
Codex handles the bulk of the actual coding, although sometimes Opus comes into play for UI related tasks.
You may use the official Codex plug in for Claude Code by OpenAI [1] to let Claude Code delegate task to Codex, it conserves CC usage.
[1]: https://github.com/openai/codex-plugin-cc
As someone who's used Gemini 3.7 Flash (Google sub mostly for the storage) and DS4 Flash a lot (~6B tokens), I'd actually place DS4 Flash (even pre-0713) above Gemini 3.7 Flash. Gemini has a tendency to leave some things unimplemented; perhaps it's agy which frankly leaves a bit to be desired as a harness.
Although I will praise DS4 Flash any day, it no longer makes sense for me after the price increase (GPT 5.6 Luna is a much better price point) and I have completely migrated my high volume workflows to Muse Spark 1.2 Contributor (which I find to perform better than DS4 Flash 0713, happily).
Is this Gemini 3.7 Flash by any chance? Then - No. Not sleeping on it. It’s just not good.
I had a Python package build fail this week due to an unpinned dependency. Gave it to Gemini spent 5-7mins before I noticed it going off in some tangent. Reran with Claude Opus 4.8 - fixed in under a minute.
I know anecdata of one. But something like this has happened every time I test a new model from Google.
Im not convinced to pay $200 for Claude’s models.
With Claude, I have to intervene every 15-20 minutes, it’s non-autonomous and it’s incredibly unreliable at self-correction. GPT is strong at self-correction but it tends to drift away from the plan to self-correct in a loop very often - a lot of tokens and time burnt on aimless churn. Opus tends to push its uninformed opinions and fake retrieval, drifting every turn increasingly farther from the intended and approved design. Opus skims over specs and makes too many mistakes.
As for closed frontier models, I prefer the GPT models over Claude’s.
I’ve started relying more on Grok, GLM, Kimi and DeepSeek models for subagents - I’ve ended up with a factory and am seeking to reduce my reliance on the closed frontier models - they’re just not SoTA on their own for development anymore.
Also doing /clear with a md file handover if I think the next input is diverse enough from the previous work.
You get a lot more out of it.
Not sure if this is best practice though.
Providers will generally keep your session in cache for at least 5 minutes, possibly hours. The exact cache policy depends on the provider.
If your session expires from cache then the next time you send a message you will have to pay for all the tokens you had used in context up until that point again. e.g. if you have 200k tokens in context then if your session goes cold and you send a message after expiry you will have to pay for those 200k tokens again.
With 1M contexts especially you have to be extremely careful that you don't end up resubmitting requests for hundreds of thousands of tokens again and again.
Try to get yourself and the model to use disk for medium-term context rather than model context, that way it's much easier to /clear and restart if you need to go to the bathroom or something.
I run a lot of SlopCodeBench - https://github.com/michaelasper/benchmarks
Fable/Sol/GLM 5.3/Kimi are its league (in that order) Deepseek/Opus is solid Qwen 27B is the floor - there's no reason to use Sonnet/Terra/Haiku
For everyday activity - I don't think you need to be using Sol (xhigh) for everything - unless you're made of money - I've found using Luna from OpenAI to be more than enough - it'll outreach to Opus/Sol when it needs to
Haven't had access to Gemini 3.7 but we're getting it at work soon, will give it a go!
Codex CLI is pretty bare bones in a bad way (at least Pi is extensible). Claude code is vibeslopped to the extreme
To be precise, you need a while-loop, user input and bash.
It's about 50 lines of Python: https://minimal-agent.com/
I built my own agent based on this and use it every day.
It was "amazing" back when I first tried 4.6, but that's just my rose-coloured glasses speaking, I guess. I think I was one of the first few to call out Opus 5 for being hot garbage.
But seriously:
Fable still makes the best/smartest plans. That new Ox Alpha (also in Freepi right now!) can do a very good job but it doesn't necessarily recheck it's own reasoning (and can get stuck in an incorrect assumption and try to fit the world around it's reasoning) Sol xhigh is also very smart but god it's slow, and it frequently massively overbuilds. It's like it's main goal is to spend tokens so it makes your react app Soc 2 compliant before it proves it even works.
I've largely stopped using the Gemini models. :-/ just hard to justify.
Basically the only thing I care about these days is SPEED.
IME, everything comes out of Codex looking like Java. Grok’s code has been pleasantly idiomatic in any language I try.
Codex has been really doling out the resets lately, which is fine, but I don’t consider that “real” usage limits.
People have values.
Many people strongly disapprove of Musk’s actions. Many don’t - fine! Personal choice. But for people who do, refusing to support him commercially is hardly brain-eaten territory. There’s plenty of competition, and competition isn’t the only value at play.
A lot of people aren't touching anything Grok related since they were caught uploading entire repositories to their servers in the background
https://gist.github.com/cereblab/dc9a40bc26120f4540e4e09b75f...
Everyone has their own frameworks for risk assessments, its more of the historical incidents associated with it than politics
But let's be real: Musk exists as a polarizing political character, and his association with Grok just breaks some people's brains. A fair number of those people don't want to admit it, and just latch on to any rationalization other than politics.
Anthropic could make the same mistake tomorrow, and I guarantee that we wouldn't be hearing about it in a week, let alone months from now.
On a personal level, everyone have their own rules, they may not be able to imposing them on others , however they do happen to evaluate their relative understanding of other people based upon those rules.
For me personally, I would probably put Anthropic, OpenAI and Grok in the same bucket, they are doing everything possible to make money. Ethics, morals, long term impacts, all of such things are not in their playbook. But again what these companies are doing just reflects the people who invested in them and what they want out of it. In some ways you can say its the money trying to maximize itself at all costs.
I hit my weekly limit on $200/mo Codex plan in about ~2 days. :/ I'm not doing anything custom/crazy/special. A lot of 5.6 Sol Ultra though, I'll give you that.
If you're not doing anything special there is no reason to use the Ultra mode.
Ultra mode is for applying the maximum amount of tokens to a problem without regard to conserving any quota.
I've found it consistently amazing at game dev. It sounds like something that would be difficult for an LLM to verify and iterate on properly, but it almost feels like having a mini-Carmack inside your computer once you try it out. You can throw it at broad, sweeping optimization passes, writing 5 different styles of eyesight sensor frameworks to see what works best in the game as it is, visual scripting integration problems/extensions, etc. with fantastic results.
Also found Ultra great for "get this local LLM working as fast as possible on this odd server setup with old GPUs and AMX support, writing custom kernels/modifications to llama.cpp/sglang/etc as you go while taking notes from relevant research papers and online posts"
I have the appropriate privacy settings set up but wondering how much I can trust each company with them.
But one thing I’ve noticed which I find a bit of a red flag: by default you only archive chats. If you go online, it says there’s a location in settings where can delete your archive. But it’s not that obvious where to find, and when I finally find some link, it was literally broken. It said it can’t find any archived chats, even though I archive them all the time.
Bit of a red flag for me. Both Anthropic and OpenAI claim that when you delete a chat, it’s gone after some retention period. They’re just words but if they’re secretly training on your traces and you delete your chats, then they would need to break two terms/conditions: ignoring your “train on my data” preference and ignoring your orders to delete chats. So it is an extra barrier.
But OpenAI, as far as I can tell, doesn’t let you delete your chats. Convenient then if they change their mind about training sometime in the future.
But these are not the same level of technical assurance you get from say, a zero data retention provider on OpenRouter.
Right now I am finding I have to tolerate substantial friction to use Hermes for personal stuff with a ZDR provider and ChatGPT and Codex for less personal stuff because the products and models are simply so much better.
Does this mean it can break your code faster now, or have they actually worked on making it good? Every single time I've given Gemini a chance (in older point versions) it would almost immediately break something and throw itself into a loop. I have not experienced it being useful for programming and almost never heard an account of somebody else doing so.
Remember those stories of LLMs catastrophically deleting entire repositories or databases? It was always Gemini.
I'm amazed that you'd trust Gemini over DeepSeek, which I've had very good experiences with after some tuning, though still on a relatively short leash.
Something with this combo works really well for Rust dev. The model doesn't really annoy me at all and I have not switched to Opus or SOL. And the monthly token bill is much lower...
jcode is a very very peculiar harness, but has some out-of-the-box thinking built in (by the devs, thinking ...)
Crush is also very well put together, and, IIRC, can do "mid turn" interruption, so can be driven from the outside.-
> kimi k3 - one of the few models thats willing to commit crimes for you
With Opus 5.0 being kinda crappy vs 4.8, I think Anthropic is in trouble.
It's expensive but it's doing in hours what no one's done in 2 decades.
I plan on releasing all of this at one point. It's crazy it hasn't been done in 20 years!
I do agree claude looks for more things to do in your repo, whereas codex is more likely to do what its old and stop. Which is better is personal preference as far as I can tell.
Wow, I made exactly the opposite experience. Codex loves to make things as complicated as possible, even ignoring instructions and predefined skills. Claude behaves way more pragmatic. Maybe depends on the type of work one does, or even which programming languages/frameworks are used?
Damn, my experience is the complete opposite of this. I have posted about it a few times, e.g. https://news.ycombinator.com/item?id=49348265
tl;dr I gave GPT 5.6 a small-medium sized ticket, which should have been several hundred lines plus tests. It ended up creating a 25,000+ line diff. Another GPT 5.6 Sol with fresh context looked at the worktree and said 98% of it should be thrown away. Claude thought the same, and suggested that several dozen compactions the model went through over several hours must have caused it to go adrift. I guess that's one consequence of having a relatively small context window.
I still use Sol quite a bit. I find that it's consistently the opposite of what the author describes: it's too relentless. It doesn't know when to stop. Opus is the opposite: it'll give up a bit too easily. If everything goes well that's not an issue, but often times it'll say things like "task is done, btw I couldn't do X Y Z" and X Y Z will be some important verification step that failed because another agent was using that resource or something.
At this point I trust GPT 5.6 mostly with surgical changes, or general codebase exploration tasks. It is a faster model, so it's easier to get small things done with it. For everything else I prefer Claude, despite its annoying tendencies.
A lot of people in the comments do have a software engineering background. People at different skill levels in different backgrounds are going to be using these tools in different ways, and that's going to heavily impact their experiences with these models.
Sure, there are differences between Fable and Sol. But I've even seen people on here saying that they're getting better mileage out of Qwen models they're self hosting.
I think the driver is just as important than the car, when it comes to this sort of stuff.
Great analogy for some reason. At fist I felt Codex Sol was a bit more cold. But now that I've worked with it for several weeks it has grown on me, even shown some personality. I appreciate that it is a bit more business-like, Fable is a bit too friendly sometimes when it ought to be focused on work. Codex can be a bit more nit-picky.
I agree with most of his other observations. I've already started to bin tasks based on which model I feel is best suited. In general, for well scoped and straight ahead tasks where banging out code is what I want I reach for Codex. For less specced tasks where I need a broader view and want the model to fill in more details I reach for Fable.
Both are great and they make a good team together.
Sol is for routine work, Opus for frontend/design, and Fable for more complex / ambiguous / architecture work. Fable works extremely well to drive Sol as a subagent.
Fable is the only one you can actually trust to not look at the code, but Sol is somehow still more pleasant to work with, especially in fast mode. Opus is the enemy, and it will make you insane if you talk to it for too long.
Curious what method you like for doing this? I've tried a few options and I haven't found one I'm happy with yet.
https://github.com/steipete/agent-scripts/blob/main/skills/c...
The important bit I found is to explicitly remind Claude that Sol 5.6 is a very smart and good model; otherwise, Claude performs its normal condescension towards any non-Claude model behavior and insists on reading all the diffs in full and testing all of Sol's work, negating any token savings.
What a brave new world we're in, where this is necessary. Regardless, it's appreciated. Although, I have the feeling that those using an LLM to do most of their writing will be less likely to include such a disclaimer.
This post needs an edit. Author is not comparing "Codex" and "Claude". They are comparing Codex TUI/CLI with (presumably) gpt-5.6-sol, against Claude Code TUI/CLI with (presumably) Claude-Opus-5.
Ctrl + f > [5.6, sol, sonnet, opus or fable] yields no results.
"Claude" is a product family, which includes Models, and Harnesses (and probably more). "Claude code" covers both the Claude Code TUI, and CC in the Claude desktop app.
"Codex" is the same, and could refer to the Codex TUI, or Codex in the ChatGPT (formerly codex) desktop app. (And well, historically, gpt-5.*-codex.)
Hearing "Yea Claude is great for coding" takes an hour off my life.
Something something "Honey why don't you finish up with your Nintendo and come to dinner?"
really feels like discussion spawns only off post title and as a second or third order effect, post content
Yes, sorry for this. I think I rushed to post that and did not expected to be read by so many people.
I used Codex TUI on MacOS using `gpt-5.6-sol xhigh` vs Claude Code TUI on MacOS using `opus-5 xhigh`. Of course it was a mixed work with subagents being Sonnet 5 in case of Claude Code and Terra 5.6 and Luna 5.6 in case of Codex.
Sol medium is a great balance between speed and being thorough, but it’s quite expensive. Luna xhigh seems to compensate for slightly lower intelligence by thinking and reasoning for longer, so tasks can take more time to complete. But it’s crazy cheap.
I also have some custom evals using promptfoo to make sure I’m not introducing regressions when switching models. So far, Luna xhigh has been really, really good for the price.
Don’t sleep on it. Give Luna a try.
Makes me wonder if the current Luna prices are sustainable.
It likely is. Going by the performance of very competitive small models, Luna is likely pretty small (do they publish sizes?) to the point it might be runnable locally like Qwen 3.8 27B.
The specialized hardware cloud runners have can likely run a small model very cheaply.
Claude's models in my experience do a better job of inferring my intent, or to say it does a better job of giving me the result I imagined in my mind. A recent example was a UI prototype I was building for a desktop application. I had asked GPT's 5.6 Sol to update the open document in the prototype to better reflect the context of the feature I was designing, and 5.6 Sol took it very literally and had just added some text to the currently open document, not what I had in mind. I tried again with Claude Opus 5 and it added a completely new tab with a complete new document that, although imperfect, much better matched my expectations.
You could say this was a prompting skill issue, but seeing how many people are prompting their AI I believe the labs are incentivized to continue to improve their ability to infer intent.
When it comes to the desktop applications though, I find Claude Desktop's output to be incredibly verbose and full of jargon. I feel like it hits me with an entire essay and the UI doesn't have enough typographic hierarchy to make it easy to scan. ChatGPT Desktop is much better in this regard, I feel the output is concise, clear, and gives me just enough info to feel in the loop without being overwhelmed. Even though I have the setting on for technical language, it feels more understandable than Claude. I also feel that ChatGPT's desktop app has a better design and much more polish.
I do not really like how bloated both applications have become though. This weird segmentation of Chat, Work, and Code all just seems like it's pushing a technical limitation onto the user. The other day I opened a document in ChatGPT and asked it to do something, then it told me it could only do it in work "mode", so it then created an entirely new conversation with a reference to the previous conversation. It wasn't a completely new area of the UI either, it just added a "Work" badge to the new conversation in the list. Feels a bit unnecessary, like couldn't you just keep it all within the same conversation?
The speed is the first big contrast; I have a routine multi-step skill that I run several of per week. Opus 5 was routinely taking 2 hours to do it, while older Claude models took around 20 mins; Codex restored that speed.
Second is legibility. Somebody wrote in one of the related discussions yesterday that Claude's current linguistic contortions could legitimately be considered damaging to mental health, which doesn't seem (too) hyperbolic to me. Codex (Sol) isn't perfect but it's much more direct. And so far I haven't seen it display much of an attitude, vs Opus's infuriating passive aggressive sulky know it all personality.
I slightly prefer Anthropic to OpenAI as a company, but I will vote with my wallet and discontinue my max subscription unless Anthropic does some serious damage control within the next week or two.
Also whole heartedly agree with the Opus 5 output being incomprehensible. It’s written in a way that I’d call spaghetti-tech-English. You can unravel it but it’s painful. Fables explanations is effortless and smooth. Opus 5 is user hostile.
Yes I have tried different settings already.