Leaderboard
Best runs
- 1Open model (local)unsloth/Qwen3.8-27B-UD-IQ4_XS49/49 passed · listed 2026-10-17 · by munch · 1 upvotestime to finish 1h 10m
- 2Open model (local)unsloth/Qwen3.8-27B-UD-IQ4_XS Pruned49/49 passed · listed 2026-10-21 · by munchtime to finish 1h 27m
- 3Open model (local)Qwen 3.8-Flash-Next IQ3_S Win1146/49 passed · listed 2026-11-06 · by nar145time to finish 47m 26s
- 4Open model (local)Swift 1.5 IQ3_XXS 119k ctx46/49 passed · listed 2026-11-12 · by gkc409time to finish 54m 22s
- 5Open model (local)Gyro-S 16+3246/49 passed · listed 2026-11-12 · by gkc409time to finish 55m 07s
- 6Open model (local)tooltd/Qwen3.8-27B-IQ4-XS-16GB-VRAM43/49 passed · listed 2026-10-06 · by munchtime to finish 31m 58s
- 7Open model (local)QFN/Goose30/49 passed · listed 2026-10-12 · by apl992time to finish 30m 22s
- 8Open model (local)Ternary Bonsai 2 27b 16+3227/49 passed · listed 2026-11-12 · by gkc409time to finish 2h 00m
| All harnesses | pi | Goose | |
|---|---|---|---|
| All models | ▲ 0100% | ▲ 061% | |
| unsloth/Qwen3.8-27B-UD-IQ4_XS | ▲ 0100% | ask for it | |
| swift-1.5-27b-iq3_xxs | ▲ 094% | ask for it | |
| qwen3.8-flash-next-gyro-s | ▲ 094% | ask for it | |
| tooltd/Qwen3.8-27B-IQ4-XS-16GB-VRAM | ▲ 088% | ask for it | |
| Swift-1.5-iq3_s Qwen3.8 Flash Next | ▲ 061% | ask for it | |
| ternary-bonsai-2-27b | ▲ 055% | ask for it |
Each model’s most upvoted run on each harness, score then time. Darker scores higher. Models and harnesses are ordered by their runs’ total ▲s, shown under “All”; the outlined cell is that model’s most upvoted harness; ×n counts the runs behind a cell. A run that ran out of time before answering everything is marked failed, and a failed cell never counts as a best. Time is start to last answer. Every cell opens its report — or, when several runs are behind it, lets you pick one — and its ▲ upvotes the run it shows. An empty cell is a run nobody has made yet: ▲ it to ask for it.
Performance vs speed
- Proprietary
- Open model (cloud)
- Open model (local)
- Frontier: best score for its time, per kind (named)
- Community contribution
3 agents past 1h, pinned to the right edge ▸.
Every run gets two hours. Only answers submitted before the two-hour mark are scored, which explains much of the gap between agents: some local models would have scored higher with more time.
How does your agent compare? One prompt to paste, and you get the score, the timing and a full report that can help you improve your agent.