Not evaluated
IMO 2025: 35 of 42 points (gold level) announced by OpenAI
Graded by three former medallists (self-announced)
The gist
OpenAI announced on July 19, 2025, in a post by Alexander Wei, that its experimental reasoning model scored 35 of 42 points at IMO 2025, a gold-medal score.
DiscoverIA status on October 11, 2026: not evaluated, meaning no third-party evaluation found.
Text written by an AI, reviewed by Antoine Olivier,
Cite this entry
A single status snapshot since Oct 11, 2026: a citation always points to a dated snapshot.
Cite snapshot 1, Oct 11, 2026
APA citation with the dated link
BibTeX @misc, key discoveria-imo-2025-gold-ai-r1
@misc{discoveria-imo-2025-gold-ai-r1,
author = {{DiscoverIA}},
title = {IMO 2025: 35 of 42 points (gold level) announced by OpenAI, graded by three former medallists (self-announced)},
howpublished = {DiscoverIA registry, entry imo-2025-gold-ai},
note = {Status “Not evaluated”, snapshot 1, recorded on October 11, 2026. Competition},
year = {2026},
url = {https://discoveria.ai/entry/imo-2025-gold-ai#releve-1},
urldate = {2026-10-11}
}Dated link opens this snapshot in the history
To understand ≈ 20 s reading time, estimated
What it changes
If the score holds, a language model would have reached gold-medal level at the international mathematics olympiad (IMO). The proofs written by the model are published, so they can be read and checked.
What remains unclear
According to Euronews, OpenAI evaluated its model on the 2025 problems and published its results before any official verification: the grading came from three former IMO medallists not named in the announcement. A Google researcher disputes the marking scheme applied to OpenAI; this is not confirmed.
In the glossary : Status: not evaluated
For experts
Sources, method, key facts, status history, report an error
Status
Not evaluated
Recorded on · euronews.com
Step 1 of 4 on the scale: not evaluated. Next step: under review.
- Announced
- Recordednot evaluated
- nextTo follow
What this status means
Source reading : Source read word for word (level 3 of 3)
The result is declared neither right nor wrong.
To move to “Under review”: a public review under way. To “Peer reviewed”: a journal publication or a review by named experts.
Sources
- Main sourcex.comlink only, reading not recorded
- Claimant's statementsimonwillison.netpage read word for word · consulted on Oct 18, 2026 ·
The model earned 35/42 points in total, enough for gold!
- Repositorygithub.compage read word for word · consulted on Oct 18, 2026 ·
This repository hosts the proofs produced by our experimental reasoning LLM during its evaluation on the 2025 International Math Olympiad.
- Status sourceeuronews.compage read word for word · consulted on Oct 18, 2026 ·
OpenAI evaluated its model on the 2025 IMO problems and self-published its results before official verification
- Corroborationtechcrunch.compage read word for word · consulted on Oct 20, 2026 ·
IMO did not respond to TechCrunch’s request for comment.
Method
According to Alexander Wei's announcement, quoted by Simon Willison, three former IMO medallists graded each proof independently, and the marks were settled unanimously. The model's proofs are in the aw31/openai-imo-2025-proofs repository. DiscoverIA status: not evaluated, meaning that no third-party evaluation has been found in the sources read. Set on October 11, 2026; source read on October 18, 2026 from euronews.com (page read, quotation checked word for word: “OpenAI evaluated its model on the 2025 IMO problems and self-published its results before official verification”).
Limits
DiscoverIA has recorded no independent evaluation as of this date.
Key facts
- Claimed by
- OpenAI (announcement by Alexander Wei)
announced on Jul 19, 2025 - Nature
- competition
- Field
- Mathematics (mathematics competition)
- Checked by
- three former IMO medallists, not named in the announcement, with no official role; no check by the IMO
- Sources
- 5 : 4 read word for word, 1 link only
- This entry
- Reviewed by Antoine Olivier, Oct 18, 2026
Status history
Only one observation so far.
- 01
First observation:Not evaluatedCurrent status
Checked on 2026-10-20: no evaluation by the IMO or by a third party found; TechCrunch (2025-07-21) reports no marking by the IMO.
Source: euronews.comread word for word
An observation is added; none is ever erased.
Report an error or suggest a source
Thank you, your message was sent.
Sending failed. Try again, or write to contact@discoveria.ai.
Is your result listed?What each status requires, correcting this entry, replying to it.