“It has not succeeded in reducing costs so far, and reducing prices would only increase costs.”
Our reading:
OpenAI has not succeeded in reducing its costs so far, and reducing its prices would only increase its costs.
Assessment: Unresolved
What happened
Does the evidence support the cause?
Outcome: Unresolved
We checked, and the evidence so far doesn’t settle it.
Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.
Evidence up to . Checked .
Why
On October 18, 2023, The Decoder reported that OpenAI had stopped developing Arrakis because of quality issues [e1]. That report does not measure company-wide spending or establish that all cost-reduction efforts failed [e1]. On July 18, 2024, OpenAI advertised GPT-4o mini's cost efficiency and lower customer prices, without providing comparable internal expenditure figures [e2]. Neither item supplies a cost history from early 2023 through July 29, 2024, or measures how price cuts affected usage and total expenditure [e1, e2]. Neither clause is settled by this evidence [e1, e2].
Caveats
- Several relevant financial reports were inaccessible and could not be used as evidence.
- The evidence does not distinguish changes in total expenditure from changes in inference unit costs.
- There is no causal measurement accounting for induced usage, efficiency changes and fixed commitments.
Evidence (2 items)
Context only: The Decoder,
stopped development due to quality issues.
Why it matters: The article reports the failure of an intended efficiency improvement, citing The Information. Failure of one project does not establish that all cost-reduction efforts failed or measure total expenditure.
- Publisher
- The Decoder
- Kind of source
- Secondary summary
- About events on
- Published
Context only: OpenAI,
our most cost-efficient small model.
Why it matters: OpenAI advertised improved cost efficiency before the post. The announcement describes API affordability and supplies no comparable internal expenditure series, so it cannot decisively contradict the financial claim.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Measures
- OpenAI, GPT-4o mini API customer pricing and advertised cost efficiency, USD dollars per million tokens, company reported, 2024-07-18 to 2024-07-18
- About events on
- Published
How we judged it
- Kind of claim
- Causal claim
- Deadline
- None given. We don’t add one.
Conditions
- so far (their words)
Where we had to interpret
- The plain contextual reading concerns OpenAI's own expenditure, rather than the prices its customers pay.
- The historical comparison covers the early-2023 baseline discussed in the post through July 29, 2024. It does not concern full-year 2024 realized expenditure.
- The wider post expressly bases its cost discussion on reporting and estimates, so the accounting basis is estimate; no GAAP or other formal basis is asserted.
- Total expenditure and cost per unit of inference are both plausible meanings of costs. They must not be substituted for one another.
- The second clause asserts a necessary causal increase in total costs following price cuts, presumably through greater usage. So far is a temporal qualifier, not a conditional antecedent.
Right if: Comparable estimates of actual internal expenditure establish no reduction through the post, and evidence linking price cuts, additional usage and costs establishes the asserted necessary increase.
Wrong if: Comparable estimates establish a reduction in the relevant internal costs, or a well-supported price-cut episode shows that total costs did not necessarily increase. Customer price reductions alone do not qualify.
We wrote this test after the outcome was known.
What would settle it
Comparable internal-cost records or well-sourced estimates from early 2023 through July 29, 2024, plus evidence isolating how price cuts affected usage and total expenditure. A documented reduction in relevant internal costs or a well-supported counterexample to the necessary cost increase could contradict the claim.
We found 48 claims in this post. See them all.
Full record
Source and dates
Where's Your Ed At, How Does OpenAI Survive?, paragraph 108.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-28. The post may have been edited after 2024-07-29.
Where it was said (1)
Said once in the posts we read.
: Where's Your Ed At, How Does OpenAI Survive?
It has not succeeded in reducing costs so far, and reducing prices would only increase costs.
- Where
- paragraph 108
- Published
- 2024-07-29
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Unresolved. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.
Held backBy us
The final check failed: the write-up said more than the evidence showed.