Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

Web browsing on Custom Tasks

57.7Score

GA

Updated 1mo ago

Evaluation Results

Method	Links
GA 2026.04		57.7	0.26
OpenClaw 2026.04		50	0.76

SOTA Paper

GA

GenericAgent: A Token-Efficient Self-Evolving LLM Agent via Contextual Information Density Maximization (V1.0)

Dataset

Custom Tasks

Follow for update

@wizwand_team Discord

Related Benchmarks

Instruction-following and procedural reasoning on SOP-Bench Sequential task management and state maintenance on Lifelong AgentBench Financial workflow task execution on RealFin benchmark Long-horizon task execution on Long-horizon complex tasks (test)Web browsing on WebCanvas

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs