IdeaPulley Arlo 1 on VCBench
IdeaPulley Arlo 1 scores F0.5 31.3 on VCBench, the venture capital benchmark from the University of Oxford and Vela Research, with 32.4% precision and 27.7% recall on the 4,500-founder private test set. That is rank 10 of 34, 2.9× the F0.5 of tier-1 VCs and 3.6× Y Combinator, at $0.10 per 1,000 founders scored.
- Rank
- #10 of 34
- F0.5
- 31.3
- Precision
- 32.4%
- Recall
- 27.7%
- Cost / 1k founders
- $0.10
Scored on the 4,500-founder private test set, mean over three folds. That is 2.9× the F0.5 of tier-1 VCs. See the full leaderboard.
Compared with the reference rows
| Entry | Precision | Recall | F0.5 | Cost / 1k |
|---|---|---|---|---|
| IdeaPulley Arlo 1 | 32.4% | 27.7% | 31.3 | $0.10 |
| Think-Reason-Learn Ensemble | 40.6% | 30.1% | 37.9 | $0.87 |
| Tier-1 VCs | 23.0% | 5.2% | 10.7 | n/a |
| Y Combinator | 14.0% | 6.9% | 8.6 | n/a |
| Random Classifier | 9.0% | 9.0% | 9.0 | $0 |
Human rows are normalized to the dataset's 9% success rate, so every entry is compared on the same base rate. Costs use list prices on 2026-09-24, except this one, which is the cost the submitter reported.
How it was scored
A mix of families, so an LLM runs at scoring time alongside a trained model. The submitter did not share the components or token counts for IdeaPulley Arlo 1, so any cost shown is the figure they reported rather than one we computed from tokens and list prices.
Frequently asked questions
- What does IdeaPulley Arlo 1 score on VCBench?
- IdeaPulley Arlo 1 scores F0.5 31.3 on VCBench, with 32.4% precision and 27.7% recall, rank 10 of 34 as of 2026-10-26.
- Does IdeaPulley Arlo 1 beat human investors at predicting founder success?
- Yes. Its F0.5 is 2.9 times that of tier-1 VCs (10.7) and 3.6 times Y Combinator (8.6), after both are normalized to the dataset's 9% base rate.
- How much does IdeaPulley Arlo 1 cost per 1,000 founders?
- About $0.10 per 1,000 founders, the figure the submitter reported rather than one computed from tokens at list prices, so it is plotted but not highlighted as best value. A mix of families, so an LLM runs at scoring time alongside a trained model. The submitter did not share the components or token counts for IdeaPulley Arlo 1, so any cost shown is the figure they reported rather than one we computed from tokens and list prices.