Customer stories
What teams build on Tigris, and why they moved: a reliable foundation for high-scale workloads.
Trusted by teams building at scale
85%
fal.ai
Storage costs cut by 85%
fal serves generative models at production scale. They ingested tens of terabytes in hours and store more than 100 TB on Tigris, and call it faster and far more reliable than anything they used before.
With no egress fees on any of it, their storage bill fell by 85%.
In their words
Why teams build on Tigris.
“Tigris team has built an exceptional product that is performant and reliable for our demanding AI/ML workloads at fal. Working closely with their eng team has saved us tremendous amounts of time.”
Generative AI“Before finding Tigris, we were planning on implementing our own data distribution and replication for our training and inference workloads. Now we just point our code to a single endpoint and Tigris handles it all. We've saved months of work, not to mention the maintenance cost and peace of mind.”
AI Cloud“On AWS we have to manage multiple S3 buckets across multiple regions which is a real hassle. With Tigris we just upload once and we don't need to think about which region the objects are in — they're just available everywhere we need them. The developer experience is much smoother.”
Security“The biggest win for us has been using Tigris to host data geographically close to our users. This has given us the ability to scale horizontally in a way we've never been able to do before, and to open up architectural approaches to serving data we weren't able to consider before.”
AICustomer stories
Read the case studies.
The Fastest Generative AI in the World, Running on Tigris
Tigris saved fal.ai 85% on their object storage costs compared to other clouds with egress fees.
How Beam Runs GPUs Anywhere with Tigris
Beam uses globally accessible object storage to keep workloads portable across GPU capacity and cloud locations.
How Agentuity Built a New Cloud for AI Agents
Agentuity uses Tigris to give AI agents durable, S3-compatible storage without making teams operate regional data infrastructure.
How Parseable Built an Observability Data Lake for AI Agents
Parseable pairs fast observability workflows with Tigris object storage for agent, LLM, and infrastructure telemetry.
Event-Driven Multimodal Ingestion with Tigris and Mixpeek
Mixpeek uses Tigris object notifications to trigger multimodal indexing pipelines as new data lands in object storage.
Own Your AI Context with Basic Memory
Basic Memory uses per-tenant Tigris buckets to store durable AI context while preserving clean isolation between users.
How LogSeam Searches 500 Million Logs per Second with Tigris
LogSeam keeps high-volume log search economical by storing large datasets on Tigris while keeping query paths fast.
Why teams choose Tigris
One storage layer for data that needs to move.
Global by default
Keep data close to compute and users without regional bucket sprawl.
Predictable pricing
No egress fees means data-heavy AI workflows are easier to model.
S3-compatible
Use the SDKs, tools, and object storage patterns teams already know.
Start building on Tigris
Free to start. No egress fees. No lock-in.
npx tigris initInstalls the CLI and connects your agent