Claim · reasoning

GPT-5.4 Nano (2026-03-17)on GPQA Diamond (Epoch's own run)

55.6 to 78.5 the range across every scoreable measurement we hold
22.9 points between the highest and the lowest
72.2 the number the board uses: the lower median, a real row with its own provenance
3 declared configurations, from 1 runner

All 3 scoreable measurements here come from ONE runner, so this is not a reconciliation and is not presented as one. What it shows is how far a single runner's number moves when the model is run differently.

Every measurement

3 rows. Best first. the board takes the LOWER MEDIAN of the scoreable measurements, which is a real row with its own interval and provenance rather than an interpolated average. a vendor self-report is shown beside the independent numbers and is never one of them; it cannot move a rank.

ScoreProvenanceConfigurationWho ran itWhere to check it
78.5 / 100
as published: 78.5%
INDEPENDENT
counts toward the ranking
gpt-5.4-nano-2026-03-17
effort high
Epoch AI
2026-04-14T15:07:20.513Z
CC-BY-4.0
NO_URL2026-08-30
benchmark leaderboard
the leaderboard for this benchmark, not a link to this row
72.2 / 100
as published: 72.2%
INDEPENDENT
counts toward the ranking
gpt-5.4-nano-2026-03-17
effort low
Epoch AI
2026-08-07T00:00:23.000Z
CC-BY-4.0
NO_URL2026-08-30
benchmark leaderboard
the leaderboard for this benchmark, not a link to this row
55.6 / 100
as published: 55.6%
INDEPENDENT
counts toward the ranking
gpt-5.4-nano-2026-03-17
effort none
Epoch AI
2026-08-07T00:01:43.000Z
CC-BY-4.0
NO_URL2026-08-30
benchmark leaderboard
the leaderboard for this benchmark, not a link to this row
← The Forge Rankings

Snapshot 2026-08-30-62e4e85f043b, manifest dc46131315046f1e, as of 2026-08-30. where only one party has measured a model on a benchmark, this page says so rather than presenting a single number as a reconciliation.