SpeedrunBench

What is SpeedrunBench?

Frontier LLM agents have been shown to be capable of solving increasingly complex tasks for which humans have measurable solutions.

Released

2026-10-06

Evaluates

Cybersecurity, Vision & 3D, cs.AI

Openness

unknown

Importer

Claire Radar

Review status

rule-auto-admitted

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2610.08076

SpeedrunBench paper

• Paper https://arxiv.org/abs/2610.08076

Which sources cite SpeedrunBench?

• SpeedrunBench

Related Cybersecurity, Vision & 3D, cs.AI benchmarks

• AARRI-Bench
• ABACUS
• ABC-Bench
• ActiveVision
• AD-Diff Bench
• AdSpark
• AdversaBench
• AE-UAV