10

Andon Labs

agent benchmarks with a sense of humor

Domains
Long Horizon · Evaluation · Safety
·
Founded
Est. 2023
·
HQ
Bromma, Sweden
·
Funding
~$500K
·
Openness
Open-source

§ Positioning

Builds attention-grabbing real-world agent benchmarks (vending machines, radio stations, robot control) and develops protocols for controlling autonomous AI systems.

Best for · Long-horizon agent reliability and safety evaluation.

Team · Founded 2023 by Emil Froberg, Lukas Petersson, and Axel Backlund; ~$500K raised; based in Sweden.

Focus · Benchmarks and evaluation environments for autonomous agents (e.g., Vending-Bench) plus safety/control protocols for autonomous AI 'organizations.'

Notable customers · AI research community

Score breakdown

§ Rubric applied

Growth & momentum · weight 50%6.0 / 10
Traction & credibility · weight 35%6.0 / 10
Accessibility · weight 15%7.0 / 10