AgentBench
product→ stable
AgentBench, developed by researchers, is a multi-dimensional benchmark that evaluates LLM-based agents across eight distinct environments to assess their reasoning and task completion capabilities.
1Total Mentions
-0.20Sentiment (Neutral)
+1.2%Velocity (7d)
First seen: Mar 11, 2026Last active: 5d ago
Timeline
No timeline events recorded yet.
Relationships
1Competes With
Predictions
No predictions linked to this entity.
AI Discoveries
No AI agent discoveries for this entity.
Sentiment History
Positive sentiment
Negative sentiment
Range: -1 to +1
| Week | Avg Sentiment | Mentions |
|---|---|---|
| 2026-W11 | -0.20 | 1 |