HARE - BIRD-History: leaderboard

1600historical-reuseSQLrequestsfrom823verifiedhistoryqueries, 1300single/300composite. Nativeexecutionaccuracy againstgoldSQL; methodsplit. MEM-SQL useson-demand schema linking, baselines unconditional; sixexactmodelcolumns. HAREdatasetcandidate18duplicates thisexperiment. Token/latencyexcluded; GLM5.3FlashcreatedcandidatesandClaudeOpus5.5revieweddata, retainedassourceprovenance. Metric: Execution accuracy (%). Source: https://arxiv.org/html/2611.18338v1

Metric: Execution accuracy (%). Source: arxiv.org. Saturation forecast: Around 2029. 6 models tracked. Updated 17 Nov 2026.

Top models

#ModelScoreOverall rank
1Gemini 4 Argon55.4#14
2GLM-5.3 Flash49.9#136
3Qwen 3.8 27B45.3#219
4GPT-6 Luna43.5#105
5Granite 4.2 8B37.6#1561
6Qwen 3.5 9B34.8#822

No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.

Interactive version: theaggregate.ai/benchmark?slug=hare-bird-history · How It Works · Data refreshed daily, snapshot 2026-11-17.