# Verified harness rankings

Verified rankings of AI coding-agent harnesses on real benchmarks.

No ranked runs recorded for this model yet.
| Benchmark | Active tasks | Upstream tasks |
| --- | --- | --- |
| Terminal-Bench 4.0.0 | 16 | 66 |
| Aider Polyglot 1.0.0 | 30 | 225 |
| Core suite 1.0.0 | 4 | 4 |


---

- Task suite version: Unofficial run on Arnesa Host
- Host epoch: 1
- Data use: Results may be quoted with attribution; the data is not licensed for redistribution.