“GPT 4.5 is hugely expensive.”
Our reading, the sentence that was checked:
GPT-4.5 is very expensive.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, OpenAI, in deep trouble, paragraph 11.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-openai-in-deep-trouble-c77544-a1
We recorded 18 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 2025-03-01. 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
- GPT 4.5
- hugely expensive
Our interpretation
- The plain reading is that using GPT-4.5 was unusually expensive compared with contemporary frontier models.
- I assess standard API input and output prices at launch. Marcus did not distinguish customer prices from OpenAI's internal operating costs.
- The target period is the launch period immediately preceding the post, February 27 through March 1, 2025.
Counts as supported if: Published standard API prices show a substantial, multiple-fold premium over contemporary frontier alternatives.
Counts as contradicted if: Comparable published API prices show GPT-4.5 priced near or below contemporary frontier alternatives.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
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
GPT-4.5's published API prices were $75 per million input tokens and $150 per million output tokens. Contemporary Claude 3.7 Sonnet prices were $3 and $15 respectively. Premiums of 25 times and 10 times support the ordinary description 'hugely expensive' under the stated API-price reading.
Limitations
- The conclusion concerns customer API prices, not OpenAI's internal costs.
- It does not establish poor value for every task or compare subscription prices.
Evidence (2 items)
Supports the claim: OpenAI,
GPT-4.5 is very large and compute-intensive, so it’s not a replacement for GPT-4o.
Why it matters here: The same paragraph lists standard input at $75 and output at $150 per million tokens. Against the contemporary Anthropic prices in e2, those are premiums of 25 times for input and 10 times for output.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Arithmetic
- Input premium = 75 / 3 = 25; output premium = 150 / 15 = 10. (inputs: OpenAI announcement: GPT-4.5 standard input $75 per million tokens; output $150 per million tokens.; e2: Claude 3.7 Sonnet standard input $3 per million tokens; output $15 per million tokens.; Exact ratios; no rounding.)
- Describes events of
- Public from
Context only: Anthropic,
Claude 3.7 Sonnet has the same price as its predecessors: $3 per million input tokens and $15 per million output tokens—which includes thinking tokens.
Why it matters here: Provides the contemporary competitor tariff used in the comparison. It is context because it measures another company's model.
- Publisher
- Anthropic
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, OpenAI, in deep trouble
GPT 4.5 is hugely expensive.
- Where
- paragraph 11
- Published
- 2025-03-01
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-openai-in-deep-trouble-c77544-a1: Current. Outcome: Supported. 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.