RNA-sequencing (RNA-seq) is currently the leading technology for genome-wide transcript quantification. While the volume of RNA-seq data is rapidly increasing, the currently publicly available RNA-seq data is provided mostly in raw form, with small portions processed non-uniformly. This is mainly because the computational demand, particularly for the alignment step, is a significant barrier for global and integrative retrospective analyses. To address this challenge, we developed all RNA-seq and ChIP-seq sample and signature search (ARCHS4), a web resource that makes the majority of previously published RNA-seq data from human and mouse freely available at the gene count level. Such uniformly processed data enables easy integration for downstream analyses. For developing the ARCHS4 resource, all available FASTQ files from RNA-seq experiments were retrieved from the Gene Expression Omnibus (GEO) and aligned using a cloud-based infrastructure. In total 137,792 samples are accessible through ARCHS4 with 72,363 mouse and 65,429 human samples. Through efficient use of cloud resources and dockerized deployment of the sequencing pipeline, the alignment cost per sample is reduced to less than one cent. ARCHS4 is updated automatically by adding newly published samples to the database as they become available. Additionally, the ARCHS4 web interface provides intuitive exploration of the processed data through querying tools, interactive visualization, and gene landing pages that provide average expression across cell lines and tissues, top co-expressed genes, and predicted biological functions and protein-protein interactions for each gene based on prior knowledge combined with co-expression. Benchmarking the quality of these predictions, co-expression correlation data created from ARCHS4 outperforms co-expression data created from other major gene expression data repositories such as GTEx and CCLE. ARCHS4 is freely accessible from: http://amp.pharm.mssm.edu/archs4.