back
405 comments
I'm working in large US corporation. And I see that I already have access to 5.6-Sol Ultra on my corporate account.

I haven't really used it yet.

2 months ago management was showing us scoreboards, praising leaders who used most tokens. Last few weeks, we're getting weekly emails, telling us that whenever we can - we should use cheaper models, and that we should watch the page which shows our tokens usage.

The craziest to me was someone saying “we are using AI in daily processes, now we need to automate”.

But of course to some asshole non-technical people it meant asking for their vibe coded bullshit to be merged into production without review and fighting about it.

That's kinda hilarious. Pretty soon they might just ask people to write code themselves.
Same here. It's insane that big, conservative tech companies would sign contracts saying "allow our engineers to use however much they want, we'll pay the bill later, no matter what it costs". In any other domain my company would insist on prior permission, and soft usage caps, and hard usage caps, and real-time tracking of actual dollar amounts (not just opaque tokens/credits, not just an after-the-fact view on a dashboard).

The AI companies' salespeople must be the greatest geniuses in the history of the world.

> 2 months ago management was showing us scoreboards, praising leaders who used most tokens.

Everyone is insane.

We are using this brand new hammer to build everything!

Please stop using this brand new hammer for glassmaking.

> 2 months ago management was showing us scoreboards, praising leaders who used most tokens. Last few weeks, we're getting weekly emails, telling us that whenever we can - we should use cheaper models, and that we should watch the page which shows our tokens usage.

GPT 5.5's double token cost was the threshold for me. These things are getting expensive quickly - the subsidized pricing can't go on forever.

Are corporate employees not allowed to use personal subscriptions?
Dog! Use it and tell us how it is! Stop with this token maxing moaning. You have access to a new powerful tool.
The leader boards based on token usage happened in our org for a month. Then we managed to convince the board that what matters is the reliable software shipped. Now we are back to DORA metrics.
Apologies for sounding like a billboard, but this is exactly why we built https://flowstate.inc/
I'm using local models in Ollama for most things and only use the paid corporate account when my local model gets stuck.
there seems to be very big misunderstanding about what the "ultra" is, so let me explain it basing on the codex source code:

it's similar to Claude code ultracode.

there is no ultra effort level implemented on the backend. it's just alias in the codex to max effort setting and single line addition to prompt to use subagents proactively. that's all

as far as we know pro models work differently. for once those are backend implementations and they probably run multiple parallel reasonings for any chunk and use some judgement model to pick best version as persistent one. but that's what I believe is most popular guess, because this is openai secret sauce.

there is still no way to use pro models from codex, or at leat so far there is no trace of it anywhere.

I wonder if it's related that that OpenAI has found a way to cut inference costs by half, according to The Information.

https://www.theinformation.com/newsletters/ai-agenda/openai-...

For context:

> Additionally, we’re introducing a new ultra mode that goes beyond the capabilities of a single agent by leveraging subagents to accelerate complex work.

https://openai.com/index/previewing-gpt-5-6-sol/

Can someone explain how this compares with Pro? I thought Pro was already something similar.

Hope this forces Anthropic to be less stingy with Fable.
Recently, I've been so eager to get new model releases in Codex. I'm hooked. I hope this accelerates development. Shows how dependant I have become to Codex.
I still don't know why OpenAI doesn't put gpt-5.5-pro in Codex. It's one hell of a model and easily parallels Fable/Mythos. Sure, it'll use up your quota much faster but that's the price some users are willing to pay for absolutely high quality responses.

I think gpt-5.5-pro runs 12x parallel gpt-5.5 agents behind the scene and uses OpenAI's secret sauce to synthesize their answers into one insanely good response.

Has anyone already tried 5.6 Sol in their day to day coding/development activities?

How does it compare to GPT-5.5?

Related:

Previewing GPT‑5.6 Sol: a next-generation model

https://news.ycombinator.com/item?id=48689028

the naming convention has reached the point where i am pretty sure the next one will just be called "GPT: The Reckoning"
Will it have similar limited access like Fable? It is an interesting timeline, as general access for Fable (without using extra credits) is coming to an end :(
This is such a sad state of the industry that a reply to a tweet is now considered like an announcement by the company.

No context, nothing, just a title and a random link to a tweet, which has a seemingly relevant response from someone who works at OpenAI I guess.

I’m really looking forward to see non contaminated benchmarks. In this space its obvious that every day is just another day of a race. Cross fingers we get better opus for lower price
Hope they release Sol on their subscription tiers. That'll make me switch away from Anthropic to OpenAI.
Will individual subscribers have access?
Competition is still needed to allow us users to make better use of these good models.
Nice! It never made sense to me that Pro Extended wasn't in the Codex app.
can't wait. been maxing Fable out, if Sol Ultra turns out to be as good and in Codex - that's a paradigm shift
Which will be the uses cases for this model?
They really try to force everyone into AI.

Next logical step: mandatory age sniffing but it can only be done if you have AI. Those not complying will be denied access to the www.

No Twitter, what’s he responding to?
when will it be available? do we know? I don't have X, not sure if the thread mentions it.
Are people excited about the capabilities of 5.6 over competing LLMs?
they better get that out fast, it will become totally meaningless when the next GLM gets there first.

more competition is always good for consumers.

Bruh when did understanding chatbots become like following pokemon? Wtf does any of this this mean. Tf is sol? Tf is ultra? Tf is codex? Tf happened to descriptive nomenclature?
Back to terrible naming from Open AI.
Gamechanger..
All these names mean squat
I can see they have inherited their poor product naming from Microslop
I’m all in on Anthropic. How good is frontier openAI models for coding and things?