# CRAB

Cross-environment benchmark framework for multimodal agents.

- Catalog URL: https://awesomecua.github.io/resources/github-camel-ai-crab/
- Original source: https://github.com/camel-ai/crab
- Kind: benchmark
- Review: curated
- Scope: core
- Environments: desktop, android
- Topics: evaluation, cross-app
- Evidence checked: 2026-10-16T06:35:58Z
- ID: github:camel-ai/crab

## Limitations

Inclusion does not establish reproducibility or endorsement.

- Record JSON: https://awesomecua.github.io/resources/github-camel-ai-crab/record.json

## Structured profile

- actions: Python-defined environment actions
- evaluation: Graph-based evaluators measure progress through task requirements.
- focus: cross-environment tasks
- observations: not recorded
- setting: interactive

### Count observations

- 120 tasks — CRAB Benchmark-v0 (paper) (reported; source: https://arxiv.org/abs/2407.01511)

CRAB is also an extensible framework. The 120-task count belongs to Benchmark-v0, not every environment that can be created with the framework.

## License declarations

- code: not recorded
- data: not recorded
- document: not recorded
- weights: not recorded

## Primary evidence

- https://raw.githubusercontent.com/camel-ai/crab/HEAD/README.md (retrieved 2026-10-16T06:35:58Z; SHA-256 7b894c5474ba78294427cfd4ef0eea8cb22d76dfa0c2e8c7f01e31ad927064e9)
- https://export.arxiv.org/api/query?id_list=2406.12814%2C2406.13719%2C2406.14056%2C2406.14250%2C2406.14314%2C2406.18082%2C2406.19263%2C2407.00993%2C2407.01476%2C2407.01489%2C2407.01511%2C2407.01558%2C2407.03037%2C2407.03913%2C2407.04346%2C2407.05291%2C2407.07061%2C2407.08850%2C2407.09018%2C2407.09295%2C2407.10956%2C2407.13032%2C2407.13063%2C2407.15673%2C2407.15711%2C2407.16741%2C2407.17490%2C2407.18901%2C2407.19056%2C2407.20183%2C2408.00203%2C2408.02544%2C2408.06327%2C2408.07199%2C2408.11824%2C2408.15978%2C2409.02453%2C2409.04566%2C2409.07429%2C2409.08264%2C2409.09354%2C2409.10741%2C2409.11295%2C2409.13711%2C2409.14337%2C2409.14818%2C2409.15441%2C2409.15637%2C2409.17140%2C2409.20566&max_results=50 (retrieved 2026-10-16T06:31:32Z; SHA-256 b9a83994d8ac36ccdb606a5219511f8bb35ad3948ade4f9f3e28d386221aedfb)