Title-like document entity extraction
Extract and reconcile owners/entities across fragmented property docs.
Provisional leader
gpt-5-2025-08-07
Current leader based on limited benchmark evidence. Treat this ranking as directional until coverage improves.
18.3%
Best benchmark score
29.7%
Confidence
All ranked models — top 3
Ranked Models
18
Evidence Quality
79%
Evidence Points
27
Top Signal
ExtractBench Paper Baselines: credit_pass_rate_pct
All Ranked Models
| Rank | Model | Score |
|---|---|---|
| 🥇 | gpt-5-2025-08-07 Strong on ExtractBench Paper Baselines credit_pass_rate_pct and LEXam Leaderboard average_score_pct | 18.3% |
| #5 | gpt-4.1-20250414 Strong on MMLongBench-Doc Leaderboard acc_score_pct and Galileo Agent Leaderboard v2 Avg AC | 16.3% |
| #15 | claude-sonnet-4 Strong on Galileo Agent Leaderboard v2 Avg AC and LanguageBench overall:mean | 13.4% |
| #18 | gemini-2.5-pro Strong on Galileo Agent Leaderboard v2 Avg AC and LEXam Leaderboard average_score_pct | 12.0% |
| #21 | Grok-4-0709 Strong on Galileo Agent Leaderboard v2 Avg AC and Galileo Agent Leaderboard v2 Avg TSQ | 10.6% |
| #25 | gpt-4o-20241120 Strong on MMLongBench-Doc Leaderboard acc_score_pct and DuckDB NSQL Leaderboard all_execution_accuracy | 10.0% |
| #26 | gemini-3-pro-preview Strong on BFCL Multi-turn Official Multi Turn Acc and LEXam Leaderboard average_score_pct | 9.8% |
| #27 | gemini-2.5-flash Strong on LanguageBench overall:mean and Galileo Agent Leaderboard v2 Avg AC | 9.7% |
| #28 | qwen-2.5-72b-instruct Strong on Galileo Agent Leaderboard v2 Avg AC and DuckDB NSQL Leaderboard all_execution_accuracy | 9.6% |
| #29 | gpt-5-mini-2025-08-07 Strong on LEXam Leaderboard average_score_pct and Vals MedQA overall_accuracy_pct | 9.5% |
| #30 | gemini-3.1-pro-preview Strong on Vals Mortgage Tax overall_accuracy_pct and Vals GPQA overall_accuracy_pct | 9.2% |
| #31 | gpt-4o Strong on JSONSchemaBench Leaderboard medium_schema_compliance_pct and DuckDB NSQL Leaderboard all_execution_accuracy | 8.7% |
| #32 | deepseek-r1 Strong on DuckDB NSQL Leaderboard all_execution_accuracy and LEXam Leaderboard average_score_pct | 8.5% |
| #33 | gpt-5.2-2025-12-11 Strong on BFCL Multi-turn Official Multi Turn Acc and FACTS Benchmark Suite facts_grounding_score_pct | 8.4% |
| #34 | o3-20250416 Strong on Aider Polyglot Leaderboard percent_correct_pct and Vals MedQA overall_accuracy_pct | 8.1% |
| #37 | grok-4-1-fast-reasoning Strong on BFCL Multi-turn Official Multi Turn Acc and Vals CorpFin v2 overall_accuracy_pct | 6.5% |
| #38 | o4-mini Strong on Aider Polyglot Leaderboard percent_correct_pct and Vals MedQA overall_accuracy_pct | 6.1% |
| #39 | gpt-4o-mini-2024-07-18 Strong on DuckDB NSQL Leaderboard all_execution_accuracy and LLM Trustworthy Leaderboard privacy | 5.8% |
Compare Models
▶Ranking diagnostics & missing models
Source lift
Ranked
18
Sources
8
Quality
Low
Vals Legal Bench
Vals Mortgage Tax
Vals MedQA
Vals LiveCodeBench
Missing frontier models
claude-opus-4-5-20251101
Thin evidence after weightingRank #10
18.6%
claude-sonnet-4.6
Thin evidence after weightingRank #11
20.0%
grok-4-1-fast-non-reasoning
Thin evidence after weightingRank #15
14.9%