855 comments
Here's GPT-6 Luna pelicans: https://tools.simonwillison.net/markdown-svg-renderer?url=ht...
And GPT-6 Sol: https://tools.simonwillison.net/markdown-svg-renderer?url=ht...
Scroll to the bottom for the GPT-6 Sol max one: https://tools.simonwillison.net/markdown-svg-renderer?url=ht...
For comparison, here are the pelicans I got for GPT-6 Astra: https://tools.simonwillison.net/markdown-svg-renderer?url=ht... - I still like the Astra Max one best.
Here's a comparison grid showing all of the GPT-6 and GPT-5.6 pelicans at all effort levels: https://static.simonwillison.net/static/2026/gpt-6-and-5.6.h...
The grid is actually really interesting, because it shows that the 5.6 family default to brighter colors than the 6 family.
Cached Read: ~6,500M
Input: ~150M
Output: ~20M
Approx $40 worth of usage across DeepSeek V4 Flash + MuseSpark Contributor 1.3. And a bit of both the GLM models. This is covered in a $10 subscription.
If I were to use Luna's API pricing:
$0.02 x 6,500 = $130
$0.20 x 150 = $30
$1.20 x 20 = $24
So $184. And this is assuming smaller coding sessions (<272K) beyond which Luna pricing doubles.
--
Cost wise, these models are nice for small stuff. Translations etc. Any model that does not provide multiple Mtoks of cached reads per cent is not very useful to me for coding workflows.
I dont know how they make money here
Snark aside, Luna 5.6 was (is) an incredible game-changer.
And info from the help page with message limits suggests the 50% price cut does not apply to the subscription, where they applied only a 1/3 price cut instead.
I'm not thrilled with this release.
Opus 5.5, which matches GPT-6 Astra performance at a cheaper price, is much more interesting.
By raising it from investors.
I had all the tabs open individually and harder to scan which model is which... otherwise keep up the great work! I like the grid view a lot. (Also the pages have no OG images set, which impacts what the link looks like shared)...
OG images will require me to move away from publishing in a Gist and linking to from a JavaScript page that loads the Gist. Probably worthwhile though.
That aside, the relevance these days is in comparing models and effort levels within the same model families - hence the comparison grids.
My results with 5.6 Sol were quite similar to 6, although I haven't tested it that much.
It wants things beyond what the mortals (us) know to reach for. It's not good at explaining itself, it doesn't show it's thinking. It's often not wrong. But the no compromises attitude can be unbearable to deal with. Especially given how little it cares about telling us.
if GPT 6 Sol is just 5.6 at half the price it will be everything i really ever wanted.
Astra was/is superior for planning type tasks. It was capable of doing seemingly magic things with rather vague/lazy instructions ("I need to be able to test this on Windows, maybe a qemu VM or something? Shrug." ... 1 hour later "yeah i built you a whole qemu + eval windows image + harness of powershell scripts + shell scripts to retrieve & verify harness.").
And for UI work -- which is not something I do a lot of but do here and there -- it was clearly superior to 5.6 Sol.
But it also feels sloppier? Somehow. And too expensive to use.
We'll see how Sol 6 is.
Otherwise I'm using 5.6 Sol for actual plan execution and review..
(I have not done anything quantitative here. For one thing, OpenAI’s billing pages and the codex-rs frontend make it pathetically difficult to get any real data. Some day I should wire up a proxy to extract actual stats.)
If I leave Astra overnight, I'll wake up with three new different projects, each of them 20% done and having nothing to do with my original goal.
1/ Usage limits: downstream of input/output cost, but resets and obscure windows and odd 20x plan / 5x plan != 4x usage math throw a wrench into it. Winner right now is Codex by a mile, especially when you factor in the fact that ChatGPT usage (even 6 Astra Pro) is essentially unmetered on the 20x plan. Always a bummer when asking if I should see a doctor about a rash means I can't code as much. It's also is a godsend if you use an MCP like oracle to automate the process of calling the Pro model on particularly tough problems, giving better planning results or deeper code analysis without burning usage.
2/ Context window in the harness. Claude Code wins on this. There used to be a toml file workaround for Codex to extend the GPT context window to 1m, but this stopped working on the plans and only on per-token billing (ETA: noname120 pointed out this is no longer the case and it can be enabled again [1]). 252k is just not enough. Codex's compaction is very good, fwiw, but it happens so frequently that even a model as powerful as Astra sometimes loses the plot on long-running tasks.
3/ Ability to use the plan outside of the official harness. Codex wins. Anthropic does shit like bills requests as extra usage if it sees a hermes.md in a commit.
I've subscription hopped a bunch, and at times I've had both, but I keep coming back to Codex because it wins on 2/3.
ETA: apparently I haven't been Keeping Up With the Altmans and new 20x signups have been disabled for a few weeks. I am grandfathered in, which makes the comparison above pretty much moot.
This hasn't been the case since around July. If you measure usage in raw api costs, Anthropic is actually giving more on $200 than OpenAI now. This includes resets. Usage allocation difference would be humiliating for codex subs were it not for resets. But fixing usage limits with resets is ugly, and they're not good for your mental well-being.
> Context window in the harness
Codex now allows 1M for subs with config params. But generally speaking, you shouldn't really be using 1M context. If you accidentally send a request with say, ~700k context already accumulated in a session which is outside cache TTL, you're paying full cost of these 700k tokens.
> I've subscription hopped a bunch
OpenAI actually has a new strategy to prevent subscription hopping after their 2-3 month-long marketing push to get claude-folks to switch over:
you can't buy a $200 sub anymore. So if you cancel, you won't be able to get back in. Hostage situation, essentially.
EDIT: re: usage limits, oh-my-pi maintainer has been tracking this - https://nitter.xitter.cc/_can1357/status/2090075496948060372
I think I’m gonna move back to a Claude plan. I could barely hit the $200 limit if I went non-stop on programming tasks.
When i swapped between a 200k Fable context into an Astra model (i was out of fable) the token usage in that context dropped to 150k or something.
Either there was a bug somewhere, or the same text got cut up very differently between providers.
The Claude TUI is just so much better though so I'm hoping Opus 5.5 is actually good and not just benchmaxxed.
OpenAI has no such nonsense. No separate meter. No five hour limits. I get to use Astra at max effort on literally every task if I want to, and even this somehow lasts me several days.
Anthropic got caught playing stupid "20x refers to the 5h limit" word games with their customers. Meanwhile, I have statistically verified that OpenAI Pro 20x = 4 * Pro 5x = 20 * Plus, exactly as advertised.
I quantified cybersecurity lockouts on my code review benchmark and they were significantly lower on OpenAI:
https://www.matheusmoreira.com/articles/code-reviewing-lone-...
My benchmark also suggests even OpenAI's Sol models can match Fable performance at a fraction of the cost.
OpenAI also used to have a ton of very nice features: unlimited chat separate from codex, allowing turns to finish even at 0% usage remaining. Sadly these got removed after abuse.
As a former Anthropic customer, OpenAI is simply the better company. There is no way around it. Good place to be while the chinese open weights models catch up. Claude is good but it doesn't make up for Anthropic's shenanigans.
While you're understandably not including the values of the $20 standard plans on both, I find the generosity of then token limits on ChatGPT plus vs Claude Pro (it's a huge difference) to be good representation of their respective attitudes towards the average user. You literally cannot use Claude pro to build real software, unless you're extremely frugal with your prompts and don't try anything even a little ambitious.
Also, Anthropic has zero models comparable to Luna.
They're both pretty horrible, but I find it difficult to find arguments for why Anthropic is worse than OpenAI, other than their doomtrolling. Which, in the grand scheme of things, doesn't even register.
Edit: forgot about the SpaceX thing.
Interestingly I would have drawn the exact opposite conclusion looking at my Claude and codex usage.
I can't get anything sustained out of codex in chatgpt plus, while I have been using Claude pro extensively and put on a lot of experimental task and features.
I ran into codex exhausting a 5h window on code review in minutes (like 3minutes) multiple times, while I could get Claude to implement 2~3 medium sized features with the same usage consumption.
(I also really dislike the usage resets in codex, they always make me feel like I use them wrong because I often just want to reset the 5h window, but they can only do both at once...)
> Also, OpenAI is just a company I'd rather support than Anthropic.
"You literally cannot use Claude pro to build real software, unless you're extremely frugal with your prompts and don't try anything even a little ambitious" - that's way past ridiculous. Even just using Fable most of the time, working on several ambitious projects, I have a hard time hitting the limit with a Max plan.
As far as I know Codex (at least the GUI) can automatically call the ChatGPT Chat models (including Astra 6 Pro), you just need to @ a ChatGPT Chat conversation from within Codex and tell it when to use it.
> There used to be a toml file workaround for Codex to extend the GPT context window to 1m, but this stopped working on the plans and only on per-token billing
Not true, it works again[1]. I confirm that it works both on 5.6 Sol and Astra 6, possibly other models too.
And re: the toml workaround, AWESOME! I appreciate you pointing these two things out, this is my highest-ROI HN comment thus far.
I'm quite puzzled about why Anthropic is so hellbent on blocking other coding agents. It's not like Claude Code has any secret sauce, right? And doesn't Anthropic make monkey off API usage, and their magic is on the model side anyway?
But to be fair, they don't really enforce the harness rule that much anymore. I guess if your harness doesn't do a lot of weird things like a lot of cache misses, or triggers some distillation attacks, or some broader Chinese fingerprints, they're tongue-in-cheek okay with you using a third party harness.
I think that is less of a factor now, and I think Anthropic have backed off some on being as strict (eg, AFAIK they never implemented the two-tier "claude -p" pricing model they were planning)
Having limitless webUI ChatGPT usage is much better user experience, though. I'll give them that.
(edit: Sol-6 is half the price, so maybe the usage limits are going to be way better.)
- For general chat and web search, occasional image editing, small coding work, document review etc. ChatGPT Plus is basically limitless and “just works” since 5.6. I’ve yet to give it some task it cannot do.
- When given sensible instructions, it hardly annoys with weird phrasing, glazing, or annoying constructs.
- The apps are very good (ignoring the initially terrible Codex app)
It’s easily my best spent $23 a month.
It is a honeymoon still, enshittification is coming, who knows how that will look like given how much more expensive to run LLMs backed user experience. Some back of the envelope calculations: 300 million US users * 20$ a month * 12 months = 72 billion $ a year. 72B$ is some spare change for AI labs. That assuming entire US will be subs which is unlikely and outside of the US there are not many rich countries consuming it, India is the next market, then Brazil and Philippines I think, not super rich counties to say the least. I believe total revenue to just pay for the capex build out by the end 2027 should be on the scale of hundreds of billions a year.
Still needs a LOT of work IMO.
Prices per 1M tokens Claude Opus 5.5 Claude Opus 5
Cache reads $0.20 $0.50
Input tokens $4 $5
Output tokens $20 $25
Cache writes $5 $6.25
Input
Output
Price reduction
GPT‑6 Sol vs. GPT‑5.6 Sol
$4 → $2
$20 → $10
50% cheaper
GPT‑6 Luna vs. GPT‑5.6 Luna
$0.20 → $0.10
$1.20 → $0.50
50% cheaper
The parent + subagent workflow has become critical for keeping the reasoning agent (parent) context-lean while also letting me chat to the main agent while work is getting done.
My main process is to use Fable to reason and then spawn Opus subagents, and I get amazing results, and I'm always looking into what the subagents are doing.
- Unbearably slow
- A token eating machine like no other
- Constantly compacting
- A model (like other GPT ones) that hides thinking traces and thinking summaries, which infuriates me
I've been in the Claude camp for a while, but the way it writes has left me with a a brick for a brain and wanted to see if Astra was as good as they say. Well, I can't know, because in the time it takes for it to actually build anything useful, I've moved to other ideas.
Unbearably, annoyingly slow. I keep thinking I must be doing something wrong.
https://developers.openai.com/api/docs/pricing?latest-pricin...
If these price changes mean that coding plans have effectively more usage then that's great, but Codex is surviving on resets from my own experience using it. I was glad to go back to Claude.
Its a great release, I will use both heavily.
Read the full thread on Hacker News →
Related stories
- Hacker News · 4 points · 6 days ago
- GPT-6.1 Sol replaces GPT-6 Sol after just 7 days, with near-Astra intelligenceartificialanalysis.aiHacker News · 47 points · about 15 hours ago
- Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price warsimonwillison.netHacker News · 2 points · 8 days ago
- GPT-6 Sol and Luna push the cost efficiency frontier by halving token costartificialanalysis.aiHacker News · 1 points · 8 days ago
- GPT-6 Sol and Luna push the cost-efficiency frontierartificialanalysis.aiHacker News · 3 points · 8 days ago
- Artifical Analysis - GPT-6.1 Sol Replaces 6 Sol After 7 Daysartificialanalysis.aiHacker News · 2 points · 1 day ago