LOCA-Bench 256K

Tests long-context agents as tool environments grow to a 256K-token environment description.

1 models

Higher is better.

Higher is better. Bar colour: blue proprietary, orange open weights, green other.

Metric: Agent tasks with controlled tool-environment growth
Authors: Weihao Zeng, Yuzhen Huang, Junxian He (2026)
Source: LOCA-bench: Benchmarking Language Agents Under Controllable Environment Growth ↗

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.