LLM Reference
active

GDPval-AA

Metric: Score (higher is better)

GDPval-AA is an Elo-style Artificial Analysis knowledge-work benchmark variant. Keep it separate from OpenAI GDPval percent scores.

Models ranked

5

tracked on this benchmark

Score band

1932.0 – 45.8

best → lowest tracked

Snapshot trend

+1739.20

May 29 → Jun 26 · 1 models

Leaderboard

Tracked models ranked by Score (higher is better).

Compare candidates
#Model variant and provenanceScore
1
Claude Fable 5
Version: GDPval-AA ELOHarness: Not recordedEvaluator: Not recordedObserved: Jun 9, 2026Confidence: Not recordedSource

Notes: DAT-5842: Reported as an unstarred Fable 5/Mythos 5 shared table row; retained as Fable 5 evidence per the Researcher handoff.

1932.0
2
Claude Opus 4.8
Version: GDPval-AA ELOHarness: Not recordedEvaluator: Not recordedObserved: May 28, 2026Confidence: Not recordedSource

Notes: Official Anthropic launch benchmark.

1890.0
3
GPT-5.5 Pro
Version: GDPval-AA ELO; Artificial Analysis independentHarness: Not recordedEvaluator: Not recordedObserved: Jun 26, 2026Confidence: Not recordedSource

Notes: Updated from secondary-source 1769 to 1785 per Artificial Analysis own leaderboard article.

1785.0
4
GPT-5.5
Version: GDPval-AA ELOHarness: Not recordedEvaluator: Not recordedObserved: Jun 9, 2026Confidence: Not recordedSource

Notes: DAT-6106: GPT-5.5 score from Anthropic's Fable 5 comparison table; keep separate from OpenAI GDPval percent rows.

1769.0
5
Step 3.7 Flash
Version: Not recordedHarness: Not recordedEvaluator: Not recordedObserved: May 29, 2026Confidence: Not recordedSource
45.8

How to read this benchmark

This benchmark scores models where higher is better. Use scores for directional filtering and shortlisting, not universal quality ranking; then validate pricing, context window, provider availability, and fit for your workload.

Trust this score when

  • There is a fresh timestamped snapshot (or multiple snapshots) for this benchmark.
  • The model list covers the same version family you can actually deploy today.
  • Top candidates overlap with your required routing and feature requirements.

Be cautious when

  • There is only one benchmark snapshot or the dataset appears stale.
  • The benchmark metric direction is opposite of your decision objective.
  • The score difference between options is narrow and likely within implementation variance.

Last reviewed: May 28, 2026

Resources