Touchstone
← Harnesses

bare

legacysuperseded

No tools, no scaffolding — one-shot answer on the task.

why it exists: The absolute floor: what a model does with zero help.

why retired: Showed the bottleneck was the toolkit, not the model — replaced by the scalpel tools.v1.1

Models
4
have run it
Solved
21
gauntlet benches

Models on this harness (gauntlet solved / attempted)

ModelSolvedBy domain
gemma-4-26b-a4b-it6/6
Coding 6/6
gpt-oss-20b6/6
Coding 6/6
qwen3.6-35b-a3b6/6
Coding 6/6
qwen3.5-4b3/6
Coding 3/6