Tigris Developer Documentation
Tigris is a globally distributed, multi-cloud object storage service with built-in support for the S3 API. It uses Dynamic Data Placement and Access-Based Rebalancing to deliver low-latency access worldwide — without the need to manage replication or caching.
Features
Globally Distributed Data
Data stored close to the users automatically ensuring low latency everywhere.
S3 Compatible API
Global and fast object storage with familiar AWS S3 tools, libraries, and extensions.
Zero Egress Fees
Free data egress ensures seamless and unrestricted access to your data whenever you need it.
Strong Consistency
Strong consistency ensures that you always get the correct version of your data.
Flexible Storage Tiers
Choose the right storage tier for your data to optimize costs and performance.
Public Buckets
Public buckets allow you to share data with anyone, anywhere.
AWS S3 SDKs
Tigris is a drop in replacement for S3 compatible storage. Use all your familiar tools and libraries by simply changing your configuration.
# Configure AWS CLI
aws configure set aws_access_key_id <access-key>
aws configure set aws_secret_access_key <secret-key>
aws configure set region auto
# List buckets
aws s3 ls --endpoint-url https://t3.storage.dev
# Create a bucket
aws s3api create-bucket --bucket <bucket-name> --endpoint-url https://t3.storage.dev
# Put an object in the bucket
aws s3api put-object --bucket <bucket-name> --key <key> --body <file> --endpoint-url https://t3.storage.dev
# Get an object from the bucket
aws s3api get-object --bucket <bucket-name> --key <key> <file> --endpoint-url https://t3.storage.dev
End-to-End Guides
Data Migration
Migrate data from an existing S3-compatible bucket to Tigris without downtime and without incuring egress costs.
Mount Tigris as a Filesystem
Use Tigris bucket as a filesystem for AI workloads and get unlimited storage and memory capacity.
Training with Big Data on SkyPilot
Build a multi-cloud compatible model training job that leverages Tigris to store training data and SkyPilot to manage compute.
Using model weights in Tigris
Enhance AI model deployment by optimizing cold starts and ensuring high availability without third-party dependencies.
Store vector embeddings in Tigris
Host embeddings and vector stores for real-time retrieval-augmented generation (RAG)
Docker Registry
Use Tigris instead of S3 to pull images as fast as possible and get around Docker Hub rate limits.