SWE-PolyBench Verified (TypeScript): leaderboard
SWE-PolyBench Verified TypeScript slice evaluates agents on resolving TypeScript GitHub issues from the verified set.
Metric: Score (%). Source: huggingface.co. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | HMigBot | 53 | |
| 2 | Atlassian Rovo Dev (2025-12-08) | 43 | |
| 3 | Prometheus-v1.2 + GPT-5 | 35 | |
| 4 | Kodah | 26 | |
| 5 | iSWE-Agent | 0 |
Interactive version: theaggregate.ai/benchmark?slug=swe-polybench-verified-typescript · How It Works · Data refreshed daily, snapshot 2026-10-20.