“We also know that OpenAI’s training and inference costs in 2024 were around $7 billion, meaning that either refining current models or building new ones is quite costly.”
Our reading, the sentence that was checked:
OpenAI's training and inference costs in 2024 were around $7 billion.
Assessment: Unresolved Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, Deep Impact, paragraph 95.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-deep-impact-3d6502-a1
We recorded 49 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2025-01-29. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
From the speaker
- OpenAI’s training and inference costs
- in 2024
- were around $7 billion
Our interpretation
- The ordinary reading is expenditure attributable to calendar 2024, not a forecast made during that year.
- The post cites press estimates, so the accounting basis is estimate.
- Around $7 billion allows ordinary approximation, but does not license substituting all research spending or annualized commitments for training and inference expenditure.
- Training may include preparatory runs and related model work; the source does not precisely delimit those costs.
Counts as supported if: Reliable retrospective estimates or records place combined calendar-2024 training and inference expenditure near $7 billion on a consistent scope.
Counts as contradicted if: Reliable retrospective evidence places that same expenditure materially away from $7 billion, with uncertainty small enough to exclude ordinary approximation.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
Contemporaneous reporting explains the $7 billion figure as expected spending. A later Epoch reconstruction reaches roughly that scale when broader research compute is included, but explicitly retains projections and an amortization assumption. This makes the figure plausible as a broad estimate without establishing the completed-year training-and-inference cost asserted in the post. The records also do not establish a sufficiently comparable contrary total.
Limitations
- Primary financial records isolating the relevant expenses were not obtained.
- The original report was inaccessible; its forecasts are represented through a readable secondary summary.
- A broader reading that includes all research compute has more support, but changes the metric's scope.
Evidence (2 items)
Context only: DeepLearning.AI,
$3 billion
Why it matters here: The article reports expected training spending and separately forecasts approximately $4 billion of inference spending. These explain the roughly $7 billion figure's origin, but are not completed-year results.
- Publisher
- DeepLearning.AI
- Kind of source
- Secondary summary
- Measures
- OpenAI, Projected calendar-2024 training expenditure, USD billions of dollars, estimate, 2024-01-01 to 2024-12-31
- Describes events of
- Public from
Context only: Epoch AI,
projections for the final quarter
Why it matters here: Epoch reconstructs roughly $7 billion of research, training and inference compute, but its inputs include final-quarter projections and an assumed research amortization schedule. It therefore offers approximate corroboration under a broader scope, not a settled realized training-and-inference total.
- Publisher
- Epoch AI
- Kind of source
- Independent measurement
- Measures
- OpenAI, Reconstructed research, training and inference cloud-compute expenditure using partly projected inputs, USD billions of dollars, estimate, 2024-01-01 to 2024-12-31
- Describes events of
- Public from
What would settle it
A retrospective 2024 expense breakdown separating training, inference and other research, with consistent treatment of amortization, credits and personnel costs.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, Deep Impact
We also know that OpenAI’s training and inference costs in 2024 were around $7 billion, meaning that either refining current models or building new ones is quite costly.
- Where
- paragraph 95
- Published
- 2025-01-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-deep-impact-3d6502-a1: Current. Outcome: Unresolved. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.