CoIR (Code Information Retrieval Benchmark) is a comprehensive and robust benchmark designed to assess the capabilities of code retrieval models. It addresses the limitations of existing benchmarks by offering a diverse set of code datasets and retrieval tasks across seven domains. CoIR comprises ten curated datasets, spanning eight distinctive retrieval tasks, including text-to-code, code-to-code, code-to-text, and hybrid code retrieval. The benchmark evaluates nine widely used retrieval models, revealing significant challenges even for state-of-the-art systems. To facilitate research, CoIR is developed as a user-friendly Python framework, installable via pip, and shares a standardized data schema with popular benchmarks like MTEB and BEIR. The paper discusses the construction of CoIR, its diverse dataset composition, and the evaluation of retrieval models, highlighting the need for specialized and adaptive code retrieval systems. The benchmark aims to stimulate further development and exploration in the field of code retrieval.CoIR (Code Information Retrieval Benchmark) is a comprehensive and robust benchmark designed to assess the capabilities of code retrieval models. It addresses the limitations of existing benchmarks by offering a diverse set of code datasets and retrieval tasks across seven domains. CoIR comprises ten curated datasets, spanning eight distinctive retrieval tasks, including text-to-code, code-to-code, code-to-text, and hybrid code retrieval. The benchmark evaluates nine widely used retrieval models, revealing significant challenges even for state-of-the-art systems. To facilitate research, CoIR is developed as a user-friendly Python framework, installable via pip, and shares a standardized data schema with popular benchmarks like MTEB and BEIR. The paper discusses the construction of CoIR, its diverse dataset composition, and the evaluation of retrieval models, highlighting the need for specialized and adaptive code retrieval systems. The benchmark aims to stimulate further development and exploration in the field of code retrieval.