Spheron review

Facts from Spheron’s own pricing and docs pages, September 2026 — 66 read from a quoted sentence on the page we archived.

Summary

Spheron rents GPUs from $0.68 per GPU-hour on demand, bills per minute, and asks no minimum commitment.

SOC 2 · partnersISO 27001 · partnersHIPAA · partnersGDPR · DPANo training on your data99.9% uptime SLAVolume discounts

From

$0.68/GPU-hr

Billing

Per minute

Minimum commitment

None on dedicated or spot

Egress

Network bandwidth included in the hourly rate

Uptime SLA

99.9%

Registered in

🇸🇬

Visit Spheron ↗

On-demand pricing

NVIDIA RTX PRO 6000 Blackwell 96GB

×1, on demand · RTX PRO 6000

$2.31/hr

Deploy

NVIDIA H200 SXM5 141GB

×1, on demand · H200

$5.92/hr

Deploy

NVIDIA L40S 48GBLowest

×1, on demand · L40S

$0.96/hr

Deploy

NVIDIA B300 SXM6 288GB

×1, on demand · B300

$10.21/hr

Deploy

NVIDIA A100 80GB

×1, on demand · A100

$1.48/hr

Deploy

See all 14 Spheron offers →

What’s good

Lowest published on-demand rate is $0.68/GPU-hr (NVIDIA RTX 5090 32GB), with no minimum commitment

Billing is per minute

No KYC or approval needed to start

Publishes a 99.9% uptime commitment, with service credits when it is missed

Commitment discounts: reserved pricing improves with duration and scale

SOC 2 Type I & II — held by data-centre partners

Plans

	Dedicated (on-demand)	Spot (on-demand)	Reserved
From	$0.72/hr (RTX 4090) Spheron · Dedicated (on-demand) Every on-demand rate we hold for Spheron, per GPU per hour, cheapest first. NVIDIA RTX 5090 32GB ×1, on-demand · RTX 5090 $0.68/hr Deploy NVIDIA RTX 4090 24GB ×1, on-demand · RTX 4090 $0.72/hr Deploy NVIDIA RTX 6000 Ada 48GB ×1, on-demand · RTX 6000 Ada $0.78/hr Deploy NVIDIA L40S 48GB ×1, on-demand · L40S $0.96/hr Deploy NVIDIA A100 80GB ×1, on-demand · A100 $1.48/hr Deploy NVIDIA RTX PRO 6000 Blackwell 96GB ×1, on-demand · RTX PRO 6000 $2.31/hr Deploy NVIDIA GH200 96GB ×1, on-demand · GH200 $3.02/hr Deploy NVIDIA H200 SXM5 141GB ×1, on-demand · H200 $5.92/hr Deploy NVIDIA B300 SXM6 288GB ×1, on-demand · B300 $10.21/hr Deploy Prices as last read from Spheron’s own pages. Open a card for every provider renting it.	Up to 50% off dedicated Spheron · Spot (on-demand) Every spot rate we hold for Spheron, per GPU per hour, cheapest first. NVIDIA RTX PRO 6000 Blackwell 96GB ×1, spot · RTX PRO 6000 $1.20/hr Deploy NVIDIA L40S 48GB ×1, spot · L40S $1.30/hr Deploy NVIDIA H200 SXM5 141GB ×1, spot · H200 $3.36/hr Deploy Prices as last read from Spheron’s own pages. Open a card for every provider renting it.	Custom
Minimum term	None	None	1 week to 12+ months; typically from 8 GPUs
Billing	Per minute	Per minute	Negotiated quote
Uptime SLA	99.99% SLA (partner-provided)	None — no SLA	Partner SLA, as displayed on the listing
Tenancy	Whole VM or bare-metal node, non-interruptible	Same hardware, interruptible without notice	Custom clusters, 8 to 512+ GPUs
Provisioning	Under 2 minutes	Same as dedicated; no grace period on reclaim	Quote in 24–48 hours
Payment	Prepaid credits (USD)	Prepaid credits (USD)	Per quote; enterprise invoicing available

Details

Billing & access

How you pay

Prepaid credits, USD; enterprise invoicing on request

Currency

USD; Singapore GST where applicable

Commitment discounts

Reserved pricing improves with duration and scale

Approval or KYC

None for standard deployments

Refunds

Credits non-refundable; requests over $10,000 considered within 30 days

Unused credits do not expire while the account is open; they are forfeited on account termination.

When funds run out

Suspension or termination without notice

Late payment

Instances suspended when the credit balance runs out

Provisioning guarantee

On-demand as available; reserved capacity guaranteed

What you get

GPUs listed

B300, B200, H200, H100, GH200, A100, RTX PRO 6000, RTX 5090, RTX 6000 Ada, L40S, RTX 4090

GB300, GB200 and R100 (Rubin, ships H2 2026) are reserve-only.

Regions

North America, Europe, Asia Pacific

Capacity per region shifts with demand; the dashboard shows current availability.

Node configuration

Up to 8× per node with NVLink

Interconnect

NVLink to 8× per node; InfiniBand 400 Gb/s on multi-node (partner dependent)

Datacenter tier

Tier 3 / Tier 4 partners

Local storage

NVMe SSD included, sized per listing

Public endpoints

Dedicated public IP; root SSH

Images

Ubuntu 22.04 with CUDA; Docker + NVIDIA Container Toolkit

Access methods

Dashboard, SSH (root), CLI

Model deployment

Self-managed — any container on your instance

FAQ

How much does Spheron cost?

As of September 2026, the lowest published on-demand rate is $0.68 per GPU-hour for the NVIDIA RTX 5090 32GB. Spheron publishes an on-demand rate for 9 GPU models. Billing is per minute.

Does Spheron have an uptime SLA?

Spheron states 99.9%. Service credits: partner's responsibility; none for spot.

Is there a minimum commitment on Spheron?

No. Spheron states none on dedicated or spot.

Does Spheron charge for egress?

Spheron states its egress terms as network bandwidth included in the hourly rate.

Can I sign up to Spheron without approval?

Spheron states that signup is self-serve. On approval and identity checks it states: none for standard deployments.

Are Spheron payments refundable?

Spheron states: credits non-refundable; requests over $10,000 considered within 30 days.

Where are Spheron's data centres?

Spheron lists north America, Europe, Asia Pacific.

Who do you contract with at Spheron?

The contracting entity is Beta360 Pte Ltd (Republic of Singapore).

Reviewed September 2026, from Spheron’s own site. 66 facts, each quoted from the page it was read on.