Skip to content
Longterm Wiki has not been actively maintained since May 2026. Some data is still updated occasionally, but please don't rely on this site being current or accurate.
Longterm Wiki

o3 scores 87.5% on ARC-AGI

resource

Metadata

Source Tableresources
Source ID457fa3b0b79d8812
DescriptionFrançois Chollet reports that OpenAI's o3 model scored 87.5% on the ARC-AGI-1 Semi-Private Evaluation set using high compute (1024 samples), and 75.7% under the $10k budget constraint, representing a dramatic step-function improvement over previous AI systems. This result challenges prior intuitions…
Source URLarcprize.org/blog/oai-o3-pub-breakthrough
Children—
CreatedFeb 21, 2026, 3:56 PM
UpdatedOct 4, 2026, 2:03 AM
Synced—

Record Data

id457fa3b0b79d8812
urlarcprize.org/blog/oai-o3-pub-breakthrough
titleo3 scores 87.5% on ARC-AGI
typeweb
summaryFrançois Chollet reports that OpenAI's o3 model scored 87.5% on the ARC-AGI-1 Semi-Private Evaluation set using high compute (1024 samples), and 75.7% under the $10k budget constraint, representing a dramatic step-function improvement over previous AI systems. This result challenges prior intuitions…
review—
abstract—
keyPoints
[
  "o3 scored 75.7% on ARC-AGI Semi-Private Eval within $10k compute budget and 87.5% at 172x higher compute (~$456k).",
  "This is a massive jump from GPT-4o's ~5%, representing a qualitative shift in novel task adaptation ability not seen before in GPT-family models.",
  "o3 was trained on 75% of…
publicationId—
authors—
authorEntityIds—
publishedDate—
tags
[
  "capabilities",
  "evaluation",
  "agi",
  "benchmarks",
  "compute",
  "alignment",
  "deployment",
  "technical-safety"
]
localFilename—
credibilityOverride—
fetchedAt—
contentHash—
stableIdsid_i6eVuYOna8
fetchStatusok
lastFetchedAtOct 4, 2026, 2:03 AM
archiveUrl—
stance—
contextNoteLandmark announcement by ARC Prize documenting o3's surprising performance on ARC-AGI-1, widely cited in AI safety and capabilities discussions as evidence of a qualitative shift in AI reasoning abilities as of late 2024.
resourcePurposeprimary_source
resourceSubtypeblog_post
typeMetadata—
publisherEntityId—
relatedEntityIds—
enrichmentStatusenriched
enrichmentDateApr 2, 2026, 7:07 PM
importanceScore0.82
contentLifecycle—
Debug info

Thing ID: sid_i6eVuYOna8

Source Table: resources

Source ID: 457fa3b0b79d8812