AgentAI Agentsv1.0

AgentQ

by MultiOn Research · open-source · Last verified 2026-03-17

Self-improving browser agent that uses Monte Carlo Tree Search and reinforcement learning to autonomously navigate and complete web tasks. Learns from its own successes and failures, progressively improving task completion rates without additional human supervision.

https://github.com/multion-ai/agentq ↗

C—Below Average

Adoption: CQuality: B+Freshness: ACitations: BEngagement: F

Specifications

License: MIT
Pricing: open-source
Capabilities: self-improvement, mcts-planning, reinforcement-learning, web-task-execution, error-recovery
Integrations: openai, anthropic, playwright
Use Cases: web-automation-research, autonomous-browsing, benchmark-evaluation, task-learning, self-supervised-training
API Available: No
Autonomy Level: fully-autonomous
Tools Used: mcts-planner, reward-model, browser-controller, experience-replay
Skills: self-learning, exploration-planning, web-interaction
Trust Score: 64
Tags: browser-agent, self-improving, reinforcement-learning, web-automation, research
Added: 2026-03-17
Completeness: 82%

Index Score

48.6

Adoption

Quality

Freshness

Citations

Engagement

Explore the full AI ecosystem on Agents as a Service