VAST Datavs
Weka


Weka: Weka for a software-only high performance file system that runs on hardware you choose

All-flash storage designed to make hard disk tiers unnecessary
As of 1 September 2026, VAST Data's pricing is not published; the vendor quotes on request. A flash storage platform whose architecture separates stateless compute nodes from shared NVMe enclosures, aiming to make all-flash cheap enough to remove the disk archive tier entirely. Softwr lists it under File Storage. VAST Data is available on Linux.
Overview
VAST Data sells a disaggregated shared-everything storage architecture in which stateless protocol servers talk over NVMe fabric to enclosures of low cost QLC flash fronted by storage class memory. Data reduction combines global deduplication with a similarity-based compression technique, and the design allows very wide erasure coding stripes that reduce parity overhead. The system presents NFS, SMB and S3 and adds a database and container layer in the newer platform releases. The distinguishing claim is economic rather than architectural: that flash at this data reduction ratio costs less per usable terabyte than a disk tier once power, cooling and rack space are counted, so the tiering problem disappears. For AI and machine learning training, where GPU clusters idle waiting on data from a slow tier, that argument decides purchases. GPU idle time is expensive enough that storage cost per terabyte stops being the deciding number. Buyers are AI and research organisations, financial services quantitative teams, life sciences and large enterprises with heavy unstructured data. The trade-off is commitment. This is a large capital or multi-year subscription purchase with no small entry point, and the data reduction ratios that make the economics work vary by dataset; a workload of already-compressed media will not reduce like a corpus of scientific data.
The honest half
Concrete and checkable, so you can decide whether any of them matter to you. This is the half of a review a vendor will not write about VAST Data.
Cross-shopped
Each pairing was judged by two reviewers asking whether a buyer would genuinely weigh the two against each other. The ones that failed were deleted rather than published.


Weka: Weka for a software-only high performance file system that runs on hardware you choose


Ceph: Ceph where cost and hardware freedom outweigh the need for extreme throughput


MinIO: MinIO if the requirement is S3 object storage rather than a high performance file system
Pricing
Taken from the vendor's own pricing page. Prices move, so check before you buy.
VAST Data Platform
On request
Capabilities
Disaggregated architecture
Stateless compute nodes separated from shared NVMe storage enclosures
Similarity-based reduction
Global data reduction beyond conventional deduplication and compression
All-flash single tier
QLC flash intended to remove the need for a separate archive tier
Multi-protocol access
NFS, SMB and S3 against the same data
Wide erasure coding
Very wide stripes lowering parity overhead at scale
GPU-oriented throughput
Designed to keep large GPU training clusters fed
Snapshots and replication
Point-in-time copies and asynchronous replication between systems
Answered, with sources
Each answer names the page it came from, so you can check it rather than take our word for it.
That is the central claim, and it holds when your data reduces well. Run a reduction assessment on your actual dataset before accepting the economics.
On usable capacity after data reduction, quoted rather than published. That makes the reduction ratio assumption the most important number in the contract.
No. There is a minimum cluster size and the entry cost reflects it. Small footprints are better served elsewhere.
Keep looking
Open source distributed storage providing object, block and file from one cluster
S3-compatible object storage server you run yourself, now under AGPL with a stripped community edition
Cloud file storage that keeps the authoritative copy in object storage and caches it at each site
Petabyte-scale object storage software on your own hardware
S3-compatible object storage distributed in encrypted pieces across independently operated nodes
Modern data protection for enterprise workloads
Open-source ZFS storage operating system for network-attached storage
Global file system putting one namespace over object storage across sites
Globally distributed S3-compatible object storage with zero egress fees
Store, sync, and share sensitive files with end-to-end encryption and zero-knowledge
S3-compatible on-premises object storage with object lock immutability
Google's consumer cloud storage app, sharing the same 15GB quota as Gmail and Google Photos
Softwr does not host reviews and shows no star rating for VAST Data, because a rating we did not collect is not ours to publish. What is here is the pricing and platform detail from the vendor’s own pages, limitations we could state concretely, and alternatives a reviewer confirmed people weigh against it. Tell us if any of it is wrong.
What people switch to, and what they give up
Every tier, and where the cost actually lands
Put it head to head with anything we hold
Its rating, and an embed for your own site
Cloud filesystem that streams large media files on demand instead of syncing them
Per member per month plus storageGoogle's consumer cloud storage app, sharing the same 15GB quota as Gmail and Google Photos
Free tier plus paid storage add-onS3-compatible object storage server you run yourself, now under AGPL with a stripped community edition
Open source, no licence feeOpen source distributed storage providing object, block and file from one cluster
Open source, no licence fee