# OpenAI is deadass using mobile game microtransactions tactics with their GPT credit system lmfao — retained record

> Retained reddit record from reddit:OpenAI.

Canonical: https://highsignal.app/data/records/c1bb4b08e70bc18f

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:OpenAI)  published 2026-11-01 13:57 ingested 2026-11-02 02:39

[original source ↗](https://www.reddit.com/r/OpenAI/comments/1xl3p3s/openai_is_deadass_using_mobile_game/)

## retained text



```
As some of you might have already seen, one of the people at OpenAI recently mentioned how users were getting "confused" between Chat and Work, with the implication that they'll merge the usage allowances too. (Are these poor confused users in the room with us right now?)

Yep... Has nothing to do with their newly added Credits System... Nothing at all... yeep. Just caring for the poor struggling users who can't tell the difference between "Do you wanna just chat or do you want to work?"

Oh right, I still gotta address what I mean by the title.

Ahem, in case you haven't checked it out yet, you can now use credits to continue using eligible features after exhausting your included usage limits. You can buy credits (V-Bucks) in bundles that expire after 12 months, with different models costing different amounts of credits per million tokens.

To figure out how much you're actually spending per million tokens, you have to, of course, convert dollars into credits, calculate the token usage, and then convert it all back into dollars.

The funny part? Their API already lists prices directly in dollars. And the credit conversions I've checked work out to the same prices. (Around 4 cents per "Credit" (V-Buck) at the standard rate.)

OpenAI is deadass giving customers the mobile game microtransactions treatment. Prepaid virtual currency, bulk purchase discounts, expiring balances, the whole package.

All this to supposedly make things less confusing for the average user.

As someone who uses API wrappers and runs smaller local models... Yeah, OpenAI... I'm suuuure "250 Credits per million input TOKENS for Astra" makes perfect sense to casual users who barely know what tokens are, let alone that there are different models beyond just "ChatGPT".

Yep... Certainly easier for them to comprehend the cost than simply stating it costs $10 per million input tokens.

Oh wait... if I recall correctly... Something something psychology trick that can get people to spend more if you abstract currency from real money... oh well...

**===Ignore If you Don't Care About the Numbers===**  

**Numbers for Nerds** (Ignoring caching and bulk discounts for simplicity... But then again, microtransactions units amirite? Always some sort of "99% SALE IF YOU BUY THE 1000 DOLARS PACKAGE!!!")

Using the standard $20 = 500 credits bundle:

\* $20 / 500 = $0.04 per credit

\* $1 = 25 credits

**GPT-5.6 Luna**

\* Input: 5 credits per 1M tokens × $0.04 = $0.20

\* Output: 30 credits per 1M tokens × $0.04 = $1.20

\* API prices: $0.20 input / $1.20 output

**GPT-6 Astra**

\* Input: 250 credits per 1M tokens × $0.04 = $10

\* Output: 1,250 credits per 1M tokens × $0.04 = $50

\* API prices: $10 input / $50 output

So yeah, in these examples, the standard credit pricing works out to the same dollar amounts as the API.

They literally just converted the dollar prices into another currency that you have to buy first lmao.
```



document fetched 2026-11-02 00:27 · canonical https://reddit.com/r/OpenAI/comments/1xl3p3s/openai_is_deadass_using_mobile_game
