# Terminal-Bench

> Terminal-Bench is a benchmark of hard, realistic tasks that an AI agent must complete by running commands in a containerized Linux terminal, graded by automated tests of the final state of the environment.

- Identifier: PTL-0203
- Category: Benchmarks
- Canonical URL: https://protologue.com/t/terminal-bench/

## Description

Grading checks outcomes, not the commands used, so agents can solve tasks any way that works. It measures the command-line skills that coding and DevOps agents rely on, such as building software, configuring systems, and debugging.

## Broader terms

- [Benchmark](https://protologue.com/t/benchmark/)

## Related terms

- [Coding Agent](https://protologue.com/t/coding-agent/)
- [SWE-bench](https://protologue.com/t/swe-bench/)
- [Code Execution Tool](https://protologue.com/t/code-execution-tool/)

## Sources

- Merrill et al. (2026). Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces. https://arxiv.org/abs/2601.11868

## Cite this entry

Protologue. (2026). Terminal-Bench. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.4.0, PTL-0203). https://protologue.com/t/terminal-bench/

License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)