pip<\/code>. AWS now recommends AWS CLI v2<\/strong> \u2014 a self-contained installer that doesn’t depend on Python at all, receives ongoing feature updates, and is the version AWS documents by default going forward. The steps below use the current v2 installation method.<\/p>\n\n\n\nPrerequisites<\/h2>\n\n\n\n\n- A Linux, macOS, or Windows machine with terminal access<\/li>\n\n\n\n
- An AWS account with an IAM user (or role) that has S3 permissions<\/li>\n\n\n\n
- That user’s Access Key ID and Secret Access Key<\/li>\n<\/ul>\n\n\n\n
Architecture<\/h2>\n\n\n\n\n- The AWS CLI<\/strong> is a client \u2014 it doesn’t store anything itself, it just sends authenticated API requests to AWS<\/li>\n\n\n\n
- Your credentials<\/strong> (Access Key ID + Secret Access Key) authenticate every request<\/li>\n\n\n\n
- Each command maps to an underlying S3 API call<\/strong> \u2014
aws s3 cp<\/code> uploads\/downloads objects, aws s3 ls<\/code> lists them, aws s3 sync<\/code> reconciles a local folder and a bucket<\/li>\n<\/ul>\n\n\n\n![\"\"](\"https:\/\/pheonixsolutions.com\/blog\/wp-content\/uploads\/2018\/10\/ChatGPT-Image-Sep-27-2026-04_43_11-PM-1024x512.png\")
<\/a><\/figure>\n\n\n\nSteps<\/h2>\n\n\n\n1. Install AWS CLI v2<\/h3>\n\n\n\n
Linux:<\/strong><\/p>\n\n\n\ncurl \"https:\/\/awscli.amazonaws.com\/awscli-exe-linux-x86_64.zip\" -o \"awscliv2.zip\"\nunzip awscliv2.zip\nsudo .\/aws\/install\n<\/pre>\n\n\n\nmacOS:<\/strong><\/p>\n\n\n\ncurl \"https:\/\/awscli.amazonaws.com\/AWSCLIV2.pkg\" -o \"AWSCLIV2.pkg\"\nsudo installer -pkg AWSCLIV2.pkg -target \/\n<\/pre>\n\n\n\n\nNote:<\/strong> The original guide installed AWS CLI v1 via Python and pip<\/code> (curl get-pip.py<\/code>, then pip install awscli<\/code>). That method still technically works, but AWS CLI v2 is now the recommended version \u2014 it’s a standalone installer with no Python dependency, and gets ongoing feature updates that v1 generally doesn’t.<\/p>\n<\/blockquote>\n\n\n\n2. Verify the Installation<\/h3>\n\n\n\naws --version\n<\/pre>\n\n\n\nYou should see output confirming an aws-cli\/2.x<\/code> version.<\/p>\n\n\n\n3. Configure Your Credentials<\/h3>\n\n\n\naws configure\n<\/pre>\n\n\n\nAWS Access Key ID [None]: <your access key>\nAWS Secret Access Key [None]: <your secret key>\nDefault region name [None]: us-east-2\nDefault output format [None]: json\n<\/pre>\n\n\n\n\nSecurity note:<\/strong> Never share or commit these keys anywhere public. If you’re working on a shared or long-lived server, consider using an IAM role instead of static keys where possible.<\/p>\n<\/blockquote>\n\n\n\n4. Create a Bucket<\/h3>\n\n\n\naws s3 mb s3:\/\/your-bucket-name\n<\/pre>\n\n\n\n5. Upload a File to the Bucket<\/h3>\n\n\n\naws s3 cp .\/tests.txt s3:\/\/your-bucket-name\/folder\/\n<\/pre>\n\n\n\n6. List Your Buckets<\/h3>\n\n\n\naws s3 ls\n<\/pre>\n\n\n\n7. Remove a Bucket<\/h3>\n\n\n\naws s3 rb s3:\/\/your-bucket-name\n<\/pre>\n\n\n\n\nIf the bucket name doesn’t match exactly, you’ll get a NoSuchBucket<\/code> error \u2014 double-check the name against your aws s3 ls<\/code> output first.<\/p>\n<\/blockquote>\n\n\n\n8. Sync a Local Folder with a Bucket<\/h3>\n\n\n\naws s3 sync .\/local-folder s3:\/\/your-bucket-name\n<\/pre>\n\n\n\nsync<\/code> reconciles the contents of a local folder and a bucket (or two buckets), only transferring files that have actually changed \u2014 more efficient than re-uploading everything with cp<\/code> each time.<\/p>\n\n\n\n9. Grant Specific Permissions on Upload (Optional)<\/h3>\n\n\n\n
cp<\/code>, mv<\/code>, and sync<\/code> support a --grants<\/code> option to set permissions on objects as they’re uploaded:<\/p>\n\n\n\naws s3 cp .\/file.txt s3:\/\/your-bucket-name\/ --grants read=uri=http:\/\/acs.amazonaws.com\/groups\/global\/AllUsers\n<\/pre>\n\n\n\nField<\/th> Options<\/th><\/tr><\/thead> Permission<\/td> read<\/code>, readacl<\/code>, writeacl<\/code>, full<\/code><\/td><\/tr>Grantee_Type<\/td> uri<\/code>, emailaddress<\/code>, id<\/code><\/td><\/tr>Grantee_ID<\/td> The actual URI, email, or canonical ID matching the type above<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\nConclusion<\/h2>\n\n\n\n
Managing S3 from the command line comes down to installing AWS CLI v2 (rather than the older pip-based v1 method), configuring your credentials once, and using a small set of commands \u2014 cp<\/code>, ls<\/code>, rb<\/code>, sync<\/code> \u2014 that map directly to S3 operations. sync<\/code> in particular is worth knowing well, since it avoids re-uploading unchanged files every time.<\/p>\n\n\n\nFrequently Asked Questions<\/h2>\n\n\n\n
Is AWS CLI v1 still usable?<\/strong> Yes, but AWS recommends v2 for new setups \u2014 it’s actively maintained with new features, while v1 is considered legacy.<\/p>\n\n\n\nDo I need Python installed for AWS CLI v2?<\/strong> No \u2014 that’s one of the main differences from v1. AWS CLI v2 is a self-contained installer with no Python dependency.<\/p>\n\n\n\nWhat’s the difference between cp and sync?<\/strong> cp<\/code> copies a specific file or folder every time you run it, re-uploading everything. sync<\/code> compares source and destination first and only transfers what’s actually changed, which is much faster for repeated backups.<\/p>\n\n\n\nRelated Articles<\/h2>\n\n\n\n
How to Create an AWS S3 Bucket Using Terraform<\/a><\/p>\n\n\n\n
![\"\"](\"https:\/\/pheonixsolutions.com\/blog\/wp-content\/uploads\/2018\/10\/ChatGPT-Image-Sep-27-2026-04_43_11-PM.png\")
<\/a><\/figure>\n\n\n\n