Pricing
Dashboard
Auto-scaling · Multi-region · Kubernetes

Elastic cloud servers that scale with your traffic.

Spin up instances in 20 seconds, auto-scale on traffic spikes, deploy across 6 regions, and pay only for what you use. Built on Kubernetes, Ceph storage, and a global private backbone. BDT and USD pricing. Hourly or monthly billing.

20-second instance boot
Auto-scaling (CPU/RAM/custom)
6 global regions
99.99% HA SLA
Managed Kubernetes
S3-compatible object storage
Terraform / Pulumi / API
Free $20 starter credit
60-sec HA failover 24/7 Bangla support bKash · Nagad · Card
cloud.sylhethost.com / global fleet
All Regions Healthy
Instances
142
vCPU Total
384
Traffic
1.2Gbps
Auto-scales
+12/h
kubectl / sylhet-cloud activity
Cluster: prod-bd-1 Region: BD-1 + 5 others Uptime: 812d 14:08:42
99.99%
HA Uptime SLA
6
Global Regions
20 sec
Boot Time
$0.008
Lowest /hour
Cloud Advantages

Why teams choose cloud servers

Cloud servers run on a highly available cluster with shared storage and automatic failover. They scale horizontally in seconds, deploy to multiple regions, and bill by the hour — ideal for variable workloads, SaaS products, and 24/7 production services.

Auto-Scaling

Horizontal + vertical

Define a policy once — 'add an instance when CPU > 70%' — and our orchestrator scales your fleet up and down automatically. Pay only for the capacity you actually use.

  • CPU / RAM / custom metrics
  • Scale-to-zero supported
  • Predictive scaling (ML)
  • Cooldown + min/max bounds

Multi-Region

6 regions, 1 API

Deploy to Sylhet, Dhaka, Singapore, Frankfurt, Ashburn, and Tokyo from a single console. Free private network between regions over our backbone. Multi-region active-active is one Terraform apply away.

  • Free inter-region bandwidth
  • Global anycast IP available
  • Region pinning for compliance
  • BDIX peering from BD regions

Pay-As-You-Grow

Hourly billing

Per-second metering with a monthly cap (you never pay more than the monthly price). Turn off instances when not in use and stop paying. Perfect for dev, test, batch jobs, and traffic spikes.

  • Per-second metering
  • Monthly price cap
  • BDT and USD billing
  • 25% off annual prepay

High Availability

60-sec failover

Your instance runs on a cluster with shared Ceph storage replicated 3x. If a host fails, your VM is automatically restarted on a healthy node within 60 seconds — no data loss, no manual intervention.

  • 3x replicated storage
  • Live migration for host upkeep
  • 99.99% uptime SLA
  • Health-checked load balancers
Cloud Plans

Choose your cloud instance — USD or BDT, hourly or monthly

Per-second metering with a monthly cap (you never pay more than the monthly price even if you run 24/7). Toggle currency and billing cycle to see live pricing. Annual prepay saves 25%.

Currency
Cycle
Save 25% with annual prepay. All plans include free private network, /29 IPv4, /48 IPv6, 100 Gbps DDoS, and Terraform access.
What's included

Every cloud account ships fully loaded

No upsell games. Every plan, from the $0.008/hr Nano to the $0.304/hr XLarge, includes the same object storage, Kubernetes, load balancers, private network, and API access.

Object Storage

S3-compatible, infinite scale, $0.02/GB/month. Perfect for backups, static assets, ML datasets, and logs. AES-256 encrypted at rest.

Managed Kubernetes

CNCF-certified K8s with managed control plane, etcd backups, version upgrades, and security patching. Just deploy your workloads.

Load Balancers

Layer 4 (TCP/UDP) and Layer 7 (HTTP/HTTPS/gRPC) with health checks, session affinity, and Let's Encrypt auto-TLS.

Private Network

Free VLAN between your instances in any region. Cross-region private traffic over our backbone at no extra charge.

Auto-Scaling

Scale by CPU, RAM, request rate, queue depth, or any custom metric. Predictive ML scaling for known traffic patterns.

Snapshots & Images

One-click snapshots of any instance. Build custom images from a running VM and deploy copies across regions.

API + IaC

REST API, official Terraform provider, Pulumi provider, CLI, and SDKs for Python, Go, JavaScript, and PHP.

DDoS Protection

Always-on 100 Gbps DDoS scrubbing on every instance. Automatic, no setup required, included in plan price.

Global Regions

Six regions, one API

Deploy where your users are. Free private network between regions over our backbone. BDIX peering from Sylhet and Dhaka keeps Bangladesh traffic sub-10ms.

BD Sylhet BD-1
Healthy
Tier-III data center in Sylhet, Bangladesh. Direct BDIX peering. Ideal for Bangladesh-facing workloads.
8 msBDIX latency
42Instances
BD Dhaka BD-2
Healthy
Tier-III data center in Dhaka, Bangladesh. BDIX + international transit. Active-active with Sylhet for HA.
6 msBDIX latency
38Instances
SG Singapore SG-1
Healthy
Equinix SG1. Low latency to South-East Asia and as a BDIX-to-global jump point. 60ms from Dhaka.
60 msFrom Dhaka
24Instances
DE Frankfurt EU-1
Healthy
Equinix FR4. Serves European users with GDPR-compliant data residency. 180ms from Dhaka.
180 msFrom Dhaka
18Instances
US Ashburn US-1
Healthy
Equinix DC11. Serves US East Coast and Latin America. 240ms from Dhaka. Major peering point.
240 msFrom Dhaka
14Instances
JP Tokyo JP-1
Healthy
Equinix TY3. Serves Japan, Korea, and East Asia. 110ms from Dhaka. Premium transit to APAC undersea cables.
110 msFrom Dhaka
6Instances
How It Works

From signup to live cloud in 5 steps

The whole process takes under 2 minutes. Pay with bKash, Nagad, Rocket, bank transfer, or international card. New accounts get $20 free credit.

1

Create account

Sign up with email + bKash/Nagad/card. Get $20 free credit, valid 60 days. No card required to start.

2

Pick a plan

Choose instance size and category (burstable / general / memory / compute). Toggle USD/BDT and hourly/monthly.

3

Choose region

Sylhet, Dhaka, Singapore, Frankfurt, Ashburn, or Tokyo. Multi-region deploy is one Terraform apply.

4

Deploy & connect

Instance boots in 20 seconds. SSH/Console access immediate. Attach load balancer, snapshot, or K8s.

5

Scale & pay

Auto-scale on traffic. Pay per second. Monthly cap means you never pay more than the monthly price.

Side-by-side

Cloud vs VPS vs Dedicated, by the numbers

Three hosting tiers, three trade-offs. Cloud wins on elasticity and HA, VPS wins on price-stability, dedicated wins on raw performance. Pick what fits your workload.

Boot time (fresh instance)
20 sec vs 60 sec vs 2 hrs
Cloud
20 sec
VPS
60 sec
Dedicated
2 hrs
Auto-scaling to 100 instances
3 min vs N/A vs N/A
Cloud
3 min
VPS
Manual
Dedicated
Manual
Failover on host failure
60 sec vs Reboot vs Manual
Cloud
60 sec
VPS
~5 min
Dedicated
4 hrs SLA
Sustained CPU performance (Geekbench 6)
2,140 vs 1,840 vs 14,820
Cloud
2,140
VPS
1,840
Dedicated
14,820
Lowest entry price (per month)
$5.84 vs $5 vs $80
Cloud (Nano)
$5.84
VPS (Starter)
$5
Dedicated (Entry)
$80
Use Cases

What teams build on SylhetHost cloud

From a single Node.js API to a multi-region Kubernetes fleet — the same cloud API scales from $0.008/hr to thousands per month.

SaaS Applications

Host multi-tenant SaaS with auto-scaling web + worker tiers, managed PostgreSQL, and per-tenant isolation. Deploy updates with zero downtime.

Most popular

Microservices

Run microservices on managed Kubernetes with service mesh (Istio), automatic mTLS, canary deployments, and per-service autoscaling.

K8s native

AI Inference

Serve LLaMA, Mistral, or your own models behind a load balancer. Scale to zero when idle, scale to 100 GPUs on a viral tweet.

GPU optional

Dev / Test / CI

Spin up disposable environments for every PR. Tear down on merge. Pay per second. Terraform applies in 30 seconds.

Ephemeral

Marketing Spikes

Eid, Black Friday, viral campaigns — auto-scale absorbs the spike, scale-to-zero keeps costs flat during quiet periods.

Eid-ready

Big Data / ETL

Spin up a 50-node Spark cluster, run the job, tear it down. Pay only for the 20 minutes of compute. Object storage for raw data.

Spark / Ray
The SylhetHost Cloud Shield

Eight layers of protection on every instance

HA, security, observability, and cost control — all included, all monitored 24/7. No paid add-ons for the basics.

100 Gbps DDoS

Always-on traffic scrubbing on every instance. Automatic, no setup. Higher tiers available for attack-prone workloads.

HA Failover

If a host fails, your instance is restarted on a healthy node within 60 seconds. 3x replicated storage, zero data loss.

Daily Snapshots

Automatic daily snapshots kept for 14 days, free. Restore any instance to any point in the last two weeks.

IAM + RBAC

Scope API tokens to specific resources, regions, or actions. SSO via Google, GitHub, GitLab, or SAML on enterprise plans.

Uptime Monitoring

Free 1-minute ping + HTTP checks from 3 regions. SMS, email, and WhatsApp alerts on any downtime or metric breach.

Cost Anomaly Alerts

Daily cost reports and instant alerts if your spend spikes 20%+ above forecast. Never get a surprise bill again.

24/7 Cloud Support

Real cloud engineers on call 24/7. Bangla or English. Average first response: 8 minutes. Free for all plans.

Compliance

GDPR-ready data residency, ISO 27001 controls, SOC 2 Type II in progress, and Bangladesh Bank data-localization compliance.

FAQ

Cloud server questions, answered

Eleven things people ask before signing up. Don't see your question? Chat with us — we answer in Bangla or English.

What is a cloud server and how is it different from a VPS?
A cloud server is a virtual machine running on a highly available cluster of physical hosts with shared storage. If a host fails, your VM is automatically restarted on another host within 60 seconds. A traditional VPS runs on a single physical host — if that host fails, your VPS goes down with it. Cloud servers also support elastic auto-scaling (add or remove instances based on load), hourly billing, and one-click multi-region deployment. VPS is simpler and cheaper for steady-state workloads; cloud is better when load varies or when high availability is mission-critical.
How does auto-scaling work?
You define a scaling policy: e.g., 'add a new Small Cloud instance when average CPU across the fleet exceeds 70% for 5 minutes; remove one when CPU drops below 30% for 10 minutes.' Our orchestrator (built on Kubernetes + KEDA) monitors your fleet every 15 seconds and spins up or tears down instances to match. You can scale by CPU, RAM, request rate, queue depth, or any custom metric you publish.
Can I pay in BDT (Bangladeshi Taka)?
Yes. SylhetHost is a Bangladeshi company — every cloud plan is priced in both BDT and USD. Hourly billing uses a monthly cap (you never pay more than the monthly price even if you run 24/7). You can pay with bKash, Nagad, Rocket, bank transfer (BDT), or international card / PayPal (USD). Annual prepay gets 25% off the monthly cap.
How fast do cloud instances boot?
Most cloud instances boot in 20-45 seconds. The cluster maintains a pool of pre-warmed VM images, so when you request a new instance, it's already 80% of the way there. Auto-scaled instances typically come online 30-60 seconds after the trigger fires. Windows and custom-image instances take a little longer (60-120 seconds).
What happens if a physical host fails?
Your cloud instance is automatically restarted on a healthy host within 60 seconds. The cluster uses shared Ceph storage, so your data is replicated 3x across different physical nodes — no data loss. We monitor host health every 5 seconds and drain workloads off any host showing signs of failure (CPU errors, ECC corrections, disk SMART warnings) before it actually fails.
Do you offer managed Kubernetes?
Yes — SylhetHost Managed Kubernetes is a CNCF-certified distribution built on upstream Kubernetes. We handle control plane management, etcd backups, version upgrades, security patching, and monitoring. You just deploy your workloads. Pricing is $0.10/hour per cluster plus the cost of worker nodes. Or use our free self-managed option where you install kubeadm yourself.
Can I deploy to multiple regions?
Yes. Pick from 6 regions at deploy time: Sylhet (BD-1), Dhaka (BD-2), Singapore (SG-1), Frankfurt (EU-1), Ashburn (US-1), Tokyo (JP-1). Multi-region deployments get free private network between regions over our backbone. Use Terraform, our CLI, or the web console to manage instances across regions from one place.
Is there a free tier?
Yes — every new account gets $20 free credit (valid 60 days), enough to run a Nano Cloud instance 24/7 for ~3 months. The Nano Cloud plan also includes 1 GB of object storage and 100 GB of outbound bandwidth per month free, forever. After the free credit runs out, you only pay for what you use.
What storage do cloud servers use?
Three options: (1) NVMe block storage (default) — fast, replicated 3x across nodes, included in plan price; (2) Object storage — S3-compatible, infinite scale, $0.02/GB/month, perfect for backups and static assets; (3) Cold storage — $0.005/GB/month for archival, slower retrieval. All three are encrypted at rest with AES-256.
Do you offer load balancers?
Yes — both Layer 4 (TCP/UDP) and Layer 7 (HTTP/HTTPS/gRPC) load balancers are available. Layer 4 is $0.025/hour, Layer 7 is $0.05/hour. Both support health checks, session affinity, weighted round-robin, least-connections, and automatic SSL/TLS termination with Let's Encrypt integration.
Can I use Terraform, Pulumi, or the API?
Yes. SylhetHost Cloud has a REST API, an official Terraform provider, a Pulumi provider, a CLI (sylhet-cloud), and SDKs for Python, Go, JavaScript, and PHP. Everything you can do in the web console, you can do via API. API tokens can be scoped to specific resources, regions, or actions.

Deploy your first cloud instance in 20 seconds

Start with $20 free credit. No card required. bKash, Nagad, Rocket, bank transfer, or international card. 99.99% HA SLA with 60-second failover on every plan.

Developer tools detected

This page contains proprietary pricing logic. Please close the developer tools to continue browsing. If you believe this is an error, click continue below.