Self-Service Workspace & Vector Database Console
Create your account, pick a target cluster, set up Square billing, configure vector databases, upload & query via REST API, and monitor real-time usage. Need a step-by-step walkthrough? Read the Developer Quickstart Docs →
1Create Account
2Cluster & Plan
3Configure DB(s)
4Upload & Query APIs
5Usage Dashboard
Step 2: Target Cluster, Capacity Plan & Square Payment Setup
Pick your target cluster and capacity tier. Hard vector and byte quotas are enforced automatically by the cluster so you never incur surprise overages.
Pending Payment
Your API key (
mk_...) is automatically routed to your chosen cluster at the edge. You can also live-migrate across clusters later with zero API key changes.
Card Verification Trial
Developer Sandbox
$5 starter credit
- • 25,000 Vectors (1024-d)
- • 110 MiB Logical Cap
- • 25 QPS Rate Limit
- •
RF=2High Availability
Most Popular Starter
Starter Tier
$29 / month
- • 250,000 Vectors (1024-d)
- • 1.1 GiB Logical Storage
- • 50 QPS Sustained
- •
RF=2Replicated
Production Workloads
Growth Pro
$149 / month
- • 2,500,000 Vectors (1024-d)
- • 11 GiB Logical Storage
- • 200 QPS Sustained
- • 2 GiB Pinned Stage-1 RAM
Dedicated High-Scale
Scale Pod
$499 / month / pod
- • 10,000,000 Vectors / pod
- • 45 GiB Logical Storage
- • 500 QPS Sub-5ms P95
- • 16 GiB Pinned Stage-1 RAM
Click 1. Setup Payment via Square to open Square Hosted Checkout, then click 2. Activate & Provision Cluster Access.
Step 3: Configure Vector Databases
Create isolated vector partitions within your workspace (1024-d cosine/dot).
| Database | Dim / Metric | Routing | Action |
|---|---|---|---|
| Activate payment above to initialize your default database. | |||
Workspace API Credentials & Cluster Routing
Scoped strictly to tenant — on cluster cluster-nvme-2n.
Step 4: Upload & Query Your Vector Database
Load the 25-article Open-Source Wikipedia Dataset (with real 1024-d BGE-M3 vectors) or custom JSON via POST /v1/ingest, then run kNN + metadata search via POST /v1/search.
A. Upload Vectors into Your Database
Recommended Quickstart: Load 25 real English Wikipedia Science, AI, Storage & Space articles embedded with
1024-d BAAI/bge-m3 vectors.
B. Query Your Database (kNN + Metadata Filter)
// Click "Load Open-Source Wikipedia Dataset (25 Articles)" then "Run Search" to inspect live hits.
C. Ready-to-Run API Snippets (cURL & Python)
Step 5: Real-Time Usage, Quota & Latency Dashboard
Live telemetry from GET /v1/tenants/{tenant}/telemetry for your workspace.
Live Vectors vs. Plan Cap
0 / 250,000
0.0% of vector quota used
Logical Storage vs. Byte Cap
0.00 MiB
Cap: 1.10 GiB (RF=2 included)
Search Latency (P50 / P95)
0.0 / 0.0 ms
Avg: 0.00 ms • P99: 0.0 ms
Metered Read / Write Units
0 / 0
Rate limit: 50 QPS (0 throttled)
Search Latency Histogram (
ms)
0 queries
≤0.5
≤1
≤2.5
≤5
≤10
≤25
≤50
≤100
Active Cluster Partitions (
GET /v1/partitions)
| Scope / Partition | Routing | Live Vectors | Logical Size |
|---|---|---|---|
| No vectors ingested yet. | |||