Da King wrote: Wed Apr 23, 2025 8:40 pm
Casual Observer wrote: Tue Apr 22, 2025 10:58 pm
Who knew what a grammar stickler Da King is. (Writes process in GPT to review my posts for consistency)
Thats not grammar. Its facts. There's a huge difference between "Get the $200 tier for $20!" and "The $20 tier is just as good as the $200 tier!"
We get it man!
Dude we got your joke a page ago, that horse is already dead.
Everyone groked on the first page that not only do you have no interest in this thing you're actually derisive. I think anyone reading this thread would be ok with you denying us your razor with in this thread going forward, thanks (I count as "anyone" even if I am the ONLY JCer who gives a shit and is excited my phone now has the goddamn Star Trek Next Generation computer.
Have fun with your 20th century toys, I think ICJ has some monitors he'd ship you to fix.
So I'm gonna post in this thread mostly for my own pleasure unless there is someone here that wants to make new jokes (Jonsey for example has been creative here, bravo) and I will respond to stupid shit that I don't have time for at 50 years old. Jesus Christ Da King you're older than me I'm quite sure of that, at least Pinner is creative with his bullshit, this is JUST SAD.
So, the details of Da King's QUIBBLE:
Da King hypothesis: CO "lied" when he claimed that JCers can get access to the EXACT SAME MODEL because the pedantic fuck thought it was funny to quibble over wording between 2 sentences so let's detail the actual differences and maybe someone who'se not autistic can tell me if your experience with both sounds like it might be EXACTLY the same thing:
Both the $20 tier and the $200 tier are Chat Interfaces with the OpenAI Orion model 4.0,
- The model also access to 3 "mini" models that are more superficial but screemingly fast
- 4.5 which is a research model that they are working to sunset because its so expensive to run so they're killing its API in July though no official end date, and
- Beta of 40 with Scheduled Tasks which seems to go greyed out periodically, Chat says access comes and goes for everyone based on what kind of processes they're already running
- So FACT # both tiers $20 and $200 have the exact model access
- It is therefore Da FACT
Actually, I'd bet having to read one of tDarcos' stories that Da King doesn't even really get what the "Model" part of ai, probably thinks its an operating system or something.
So, GIVEN Da Fact that I've already evicerated his crap, wonder if anyone cares about the actual differences between the tiers?
As said $200 is a "business tier" while $20" is a "consumer tier", don't think DaKing works in IT so someone might want to explain the diff offline. If he thinks he's going to use free AI to learn thats fair but if his post claims he knows jack about what makes a model a model - AI is very good at catching it's work so at least don't copy paste without fair attribution or else you're plagerizing on a site full of people who like to write (not me I just have to)
So lets if the UNICODE from GPT is annoying enough for some assholes.

$20/month – Pro (Individual)

Access to GPT-4 (actually GPT-4-turbo)

Faster response speeds

Priority access during high load

5x context window (128k tokens)

Image, file, and web browsing support

No shared memory across users

No API credits or integrations

No workspace or admin tools

$200/month – Team / Enterprise (per seat)

Everything from Pro, plus:

Operator (agent-syle automation)

Shared files, projects, and memory across a team

Admin panel for user control and permissions

Usage analytics

Early access to experimental features (in some cases)

(Optional) API credits depending on usage tier

Contractual support (for Enterprise)
HMM, sure are 3 extra checks and all but I challenge Da King to tell me why any of you would care ??

ChatGPT (Web Interface)

Easy to use, no setup—just type and go

Built-in memory (with Pro)

Fast responses (if load is low)

Supports file uploads, browsing, vision, etc.

No guaranteed task execution

Can “pretend to work” during high load

No control over retries, timeouts, or process logging

Fragile for long or complex workflows

Flat subscription ($20/month for Pro)

OpenAI API

Direct access to GPT models via code

Guaranteed task runners (if a job starts, it finishes)

You control:
Timeout length
Max tokens
Streaming vs. one-shot
Retries and checkpoints

Can be integrated into scripts, apps, task loops, CRMs, etc.

Stable for long-running or high-integrity jobs

No built-in memory (you must supply past messages manually)

Requires setup and usage tracking

Pay-as-you-go (per token used)
Given none of the other JCers are likely to being doing things like web scraping, multi step spreadsheet analysis, building long running data loop driven processes, AT LEAST NOT DURING HIGH LOAD THAT'S GUARANTEED STACKWEED!
If my statement had been something like they're giving away free API tokens for the $200 account then you would have read that asshole.
API tokens are for real work when it really matters (not a sales guy finding himself leads), application back ends, advanced research and analysis projects that successfully run no matter how high the LOAD, even real time real world robot control using programs written in 40 and run in MINI. I can't do any of that yet but I just bought $10 of tokens to start learning that for GOD DAMN SURE.
