Back to News
CAMEL AI, Eigent AI, and collaborators•January 11, 2026
SETA
Open Source
AI
Open Source
Reinforcement Learning
Terminal Agents
Benchmarking
### TL;DR
SETA is an open-source toolkit and environment stack designed for training reinforcement learning agents in terminal environments. It offers a comprehensive solution for developing and evaluating terminal agents operating within Unix-style shells.
Key Insights & Metrics
Pricing
Free
Cost structure
Version
1.0
Current release version
Hardware
N/A
Compute requirements
Category
Open Source
Licensing model
Region
Global
Primary region
Key Features
- State-of-the-art performance on Terminal Bench 2.0 with a Claude Sonnet 4.5-based agent
- Synthetic dataset of 400 terminal tasks covering various difficulty levels
- Unified agent design compatible with both local task runs and official Terminal Bench evaluations
Discussion
0
Upvotes
0
Downvotes
0 reviews
Sign in to leave a review
Reviews
No reviews yet. Be the first to review!