Marketplace · Singapore
Spheron
Singapore-operated GPU marketplace that resells on-demand, spot and reserved GPU capacity from seven upstream providers (Verda, Sesterce, Massed Compute and Spheron-branded supply) through one prepaid account, dashboard and REST API.
GPU types23with a current offer
Current offers250259 readings on file · last 45 days
H100 from$2.21per GPU-hour, on-demand
Entry price$0.44cheapest GPU-hour listed
Countries1227 named regions
Trust Score v0.153/100
- Company & track record47
- Infrastructure control60
- Pricing & billing transparency100
- Reliability signals45
- Security & compliance0
- Customer signal—
Scored Sep 29, 2026 from the provider's own pages. Dimensions not yet measured are excluded from the denominator. Criteria ↓
Trust Score · 5 of 6 dimensions measuredThe score, criterion by criterion
Each point comes from something the provider publishes, with the page it was read on and the date. Anything not disclosed earns zero.
Company & track record
weight 15%
47 / 100
| Founding year published | not disclosedspheron.network · Sep 29, 2026 | 0 / 4 |
| Headquarters published | Beta360 Pte Ltdspheron.network · Sep 29, 2026 | 3 / 3 |
| Legal entity named | Beta360 Pte Ltdspheron.network · Sep 29, 2026 | 4 / 4 |
| Funding or ownership disclosed | not disclosedspheron.network · Sep 29, 2026 | 0 / 4 |
Infrastructure control
weight 15%
60 / 100
| Capacity model published | third-party hostsspheron.network · Sep 29, 2026 | 4 / 10 |
| Regions published | 27 regions or countriesspheron.network · Sep 29, 2026 | 5 / 5 |
Pricing & billing transparency
weight 20%
100 / 100
| Public prices without an account | yes (read on the pricing page)spheron.network · Sep 29, 2026 | 5 / 5 |
| Billing granularity published | mixedspheron.network · Sep 29, 2026 | 4 / 4 |
| Egress policy published | included (docs overview: 'Zero ingress/egress charges'; pricing FAQ: network bandwidth included in the hourly rate)spheron.network · Sep 29, 2026 | 4 / 4 |
| Stopped-instance billing published | Stop exists only on Spheron AI (some machine types and regions) and Spheron ES; the stopped rate covers retained disk plus public IP where the provider charges for it (API example: 0.0412 USD/h for a 512 GB root disk and IP). On Verda, Sesterce, Spheron MS, Spheron AM, Massed Compute, all bare metal and all spot, an instance bills the full hourly rate until destroyed.spheron.network · Sep 29, 2026 | 4 / 4 |
| Public IPv4 policy published | included: dedicated public IPv4 per instance, all ports open by defaultspheron.network · Sep 29, 2026 | 3 / 3 |
Reliability signals
weight 20%
45 / 100
| Status page | not disclosedspheron.network · Sep 29, 2026 | 0 / 6 |
| SLA published | Pass-through only: Terms say Spheron gives no SLA of its own and provider SLAs on dedicated instances go 'up to 99.5%'. Marketing pages claim 99.9% (home, pricing) and 99.99% (on-demand and spot pages); docs say 99.95% for dedicated. Spot has no SLA.spheron.network · Sep 29, 2026 | 6 / 6 |
| SLA with a numeric uptime commitment | Pass-through only: Terms say Spheron gives no SLA of its own and provider SLAs on dedicated instances go 'up to 99.5%'. Marketing pages claim 99.9% (home, pricing) and 99.99% (on-demand and spot pages); docs say 99.95% for dedicated. Spot has no SLA.spheron.network · Sep 29, 2026 | 3 / 3 |
| Uptime history published | not disclosedspheron.network · Sep 29, 2026 | 0 / 5 |
| Observed availability (30 days) | not measured yet | — |
Security & compliance
weight 15%
0 / 100
| SOC 2 | not disclosedspheron.network · Sep 29, 2026 | 0 / 8 |
| ISO 27001 | not disclosedspheron.network · Sep 29, 2026 | 0 / 4 |
| HIPAA or GDPR DPA | not disclosedspheron.network · Sep 29, 2026 | 0 / 3 |
Customer signal
weight 15%
not measured
| Aggregated public reviews | not collected yet | — |
provisional v0.1: evidence from the provider's own pages; dimensions not yet measured are excluded from the denominatorFull method →
Market report · Oct 25, 2026 · Specialised GPU clouds, hourly rental, pods and inference · #62 of the 200 to audit · prices seen on an aggregator onlyWhat the market report notes about Spheron
Our internal report lists 248 actors with their zone, segment, status and the prices it could read, dated Oct 25, 2026. Its notes are reproduced as written, in French, with their sources.
- Listed as
- Spheron
- Cross-checkfr
- RTX PRO 6000 spot 1,14 $ (20/10/2026) ; RTX 4090 0,48 $ ; L40S 0,96 $ ; publie de nombreux comparatifs (vend aussi de la capacité).
Prices by GPU · 17 GPU types with a current offerWhat Spheron charges, against the market
Cheapest current on-demand true cost here, next to the market median for the same GPU. Negative means cheaper than the market.
| GPU | Cheapest here | Market median | vs market | Current offers | Modes | |
|---|---|---|---|---|---|---|
| H100 | $2.21 | $3.17 | −30% | 32 | on-demandspot | H100 prices |
| A100 80GB | $1.19 | $1.99 | −40% | 13 | on-demand | A100 80GB prices |
| RTX PRO 6000 | $2.35 | $2.14 | +10% | 16 | on-demandspot | RTX PRO 6000 prices |
| H200 | $4.80 | $4.76 | +1% | 17 | on-demandspot | H200 prices |
| L40S | $1.80 | $1.26 | +42% | 22 | spoton-demand | L40S prices |
| B200 | $11.12 | $7.50 | +48% | 3 | spoton-demand | B200 prices |
| B300 | $12.42 | $8.96 | +39% | 4 | on-demandspot | B300 prices |
| L4 | $1.14 | $0.80 | +42% | 3 | on-demand | L4 prices |
| RTX A6000 | $0.44 | $0.54 | −18% | 7 | on-demand | RTX A6000 prices |
| RTX 4090 | $0.48 | $0.44 | +9% | 1 | on-demand | RTX 4090 prices |
| A100 40GB | $1.68 | $1.98 | −15% | 7 | on-demand | A100 40GB prices |
| L40 | $0.89 | $0.89 | −0% | 9 | on-demand | L40 prices |
| Tesla V100 16GB | $2.81 | $1.71 | +64% | 4 | on-demand | Tesla V100 16GB prices |
| MI300X | $3.59 | $2.71 | +32% | 1 | on-demand | MI300X prices |
| RTX A4000 | $0.96 | $0.19 | +419% | 3 | on-demand | RTX A4000 prices |
| RTX A5000 | $1.70 | $0.27 | +536% | 3 | on-demand | RTX A5000 prices |
| A10 | $1.55 | $1.35 | +14% | 2 | on-demand | A10 prices |
All offers · 259 readings on fileEvery Spheron offer, with its receipt
Current offers (last 45 days) ranked by true cost, all rental modes, each with its read date. Older readings stay listed below, in grey.
| GPU | Configuration | Mode | Region | Stock | List | True cost | Read | |
|---|---|---|---|---|---|---|---|---|
| RTX A6000marketplace floor | 4× RTX A6000124 vCPU · 232 GB RAM · 750 GB disk | On-demand | CANADA-1 | In stock | $0.44 | $0.44 | 21:10 UTCfresh | Receipt |
| RTX A6000 | 2× RTX A600060 vCPU · 116 GB RAM · 300 GB disk | On-demand | CANADA-1 | In stock | $0.44 | $0.44 | 21:10 UTCfresh | Receipt |
| RTX A6000 | 1× RTX A600028 vCPU · 58 GB RAM · 100 GB disk | On-demand | CANADA-1 | In stock | $0.45 | $0.45 | 21:10 UTCfresh | Receipt |
| RTX 4090marketplace floor | 8× RTX 409088 vCPU · 200 GB RAM · 7,600 GB disk | On-demand | ce32df2f:casper-usa-2 | In stock | $0.48 | $0.48 | 21:10 UTCfresh | Receipt |
| RTX A6000 | 1× RTX A600028 vCPU · 58 GB RAM · 100 GB disk | On-demand | MON1 | In stock | $0.66 | $0.66 | 21:10 UTCfresh | Receipt |
| L40marketplace floor | 1× L4028 vCPU · 58 GB RAM · 825 GB disk | On-demand | CANADA-1 | In stock | $0.89 | $0.89 | 21:10 UTCfresh | Receipt |
| RTX A4000marketplace flooroutlier · suspended | 2× RTX A400016 vCPU · 90 GB RAM · 500 GB disk | On-demand | 90e8fe74:newyork-usa-1 | In stock | $0.96 | $0.96 | 21:10 UTCfresh | Receipt |
| RTX A4000 | 4× RTX A400032 vCPU · 180 GB RAM · 1,000 GB disk | On-demand | 90e8fe74:newyork-usa-1 | In stock | $0.96 | $0.96 | 21:10 UTCfresh | Receipt |
| RTX A4000 | 1× RTX A40008 vCPU · 45 GB RAM · 250 GB disk | On-demand | 90e8fe74:newyork-usa-1 | In stock | $0.96 | $0.96 | 21:10 UTCfresh | Receipt |
| L40Smarketplace floor | 1× L40S8 vCPU · 32 GB RAM · 256 GB disk | Spot | EU North 1 | In stock | $1.05 | $1.05 | 21:10 UTCfresh | Receipt |
| L40 | 4× L40124 vCPU · 232 GB RAM · 3,300 GB disk | On-demand | CANADA-1 | In stock | $1.06 | $1.06 | 21:10 UTCfresh | Receipt |
| L40 | 4× L40124 vCPU · 232 GB RAM · 3,300 GB disk | On-demand | CANADA-1 | In stock | $1.06 | $1.06 | 21:10 UTCfresh | Receipt |
| L40 | 2× L4060 vCPU · 116 GB RAM · 1,650 GB disk | On-demand | CANADA-1 | In stock | $1.06 | $1.06 | 21:10 UTCfresh | Receipt |
| L40 | 2× L4060 vCPU · 116 GB RAM · 1,650 GB disk | On-demand | CANADA-1 | In stock | $1.06 | $1.06 | 21:10 UTCfresh | Receipt |
| L40 | 1× L4028 vCPU · 58 GB RAM · 825 GB disk | On-demand | CANADA-1 | In stock | $1.06 | $1.06 | 21:10 UTCfresh | Receipt |
| L40 | 1× L4028 vCPU · 58 GB RAM · 825 GB disk | On-demand | CANADA-1 | In stock | $1.06 | $1.06 | 21:10 UTCfresh | Receipt |
| RTX PRO 6000marketplace floor | 8× RTX PRO 6000192 vCPU · 1,744 GB RAM · 256 GB disk | Spot | UK South 2 | In stock | $1.11 | $1.11 | 21:10 UTCfresh | Receipt |
| RTX PRO 6000 | 8× RTX PRO 6000192 vCPU · 1,744 GB RAM · 256 GB disk | Spot | EU South 1 | In stock | $1.11 | $1.11 | 21:10 UTCfresh | Receipt |
| RTX PRO 6000 | 1× RTX PRO 600024 vCPU · 218 GB RAM · 256 GB disk | Spot | US Central 1 | In stock | $1.14 | $1.14 | 21:10 UTCfresh | Receipt |
| RTX PRO 6000 | 1× RTX PRO 600024 vCPU · 218 GB RAM · 256 GB disk | Spot | UK South 2 | In stock | $1.14 | $1.14 | 21:10 UTCfresh | Receipt |
| RTX PRO 6000 | 1× RTX PRO 600024 vCPU · 218 GB RAM · 256 GB disk | Spot | EU South 1 | In stock | $1.14 | $1.14 | 21:10 UTCfresh | Receipt |
| L4marketplace floor | 1× L48 vCPU · 48 GB RAM · 500 GB disk | On-demand | ad83ad6b:paris-france-5 | In stock | $1.14 | $1.14 | 21:10 UTCfresh | Receipt |
| L4 | 1× L48 vCPU · 48 GB RAM · 500 GB disk | On-demand | ad83ad6b:warsaw-poland-1 | In stock | $1.14 | $1.14 | 21:10 UTCfresh | Receipt |
| L4 | 2× L416 vCPU · 96 GB RAM · 500 GB disk | On-demand | ad83ad6b:paris-france-5 | In stock | $1.14 | $1.14 | 21:10 UTCfresh | Receipt |
| L40S | 1× L40S16 vCPU · 64 GB RAM · 256 GB disk | Spot | EU North 1 | In stock | $1.19 | $1.19 | 21:10 UTCfresh | Receipt |
147 fresh offers · the 25 cheapest shown here, all of them in the comparison tool · prices per GPU-hour · true cost includes the add-ons the provider publishesOpen in the comparison tool →
True cost & billing · read Sep 29, 2026What the hourly price does not say
Billing granularity, minimums, what a stopped machine costs, traffic, addresses, deposits and identity checks, as published by the provider.
Billing
- Billing granularity
- mixed
- Minimum charge
- 20-minute minimum runtime on every instance (60 minutes on 2x MI300X); some machine types carry a longer provider minimum shown in the deploy wizard, charged in full on early termination
- Stopped instance
- Stop exists only on Spheron AI (some machine types and regions) and Spheron ES; the stopped rate covers retained disk plus public IP where the provider charges for it (API example: 0.0412 USD/h for a 512 GB root disk and IP). On Verda, Sesterce, Spheron MS, Spheron AM, Massed Compute, all bare metal and all spot, an instance bills the full hourly rate until destroyed.
- Egress
- included (docs overview: 'Zero ingress/egress charges'; pricing FAQ: network bandwidth included in the hourly rate)
- Public IPv4
- included: dedicated public IPv4 per instance, all ports open by default
- Billing model
- prepaid credits
- Minimum deposit
- 15 USD per deposit (maximum 5,000 USD per deposit; presets 100 / 1,000 / 5,000 USD). A reserved-cluster bid holds 100 USD by default.
- Free credits
- not published
Account
- Payment
- card · crypto (yes: USDC and USDT stablecoins via CopperX (settled on-chain; no Stripe invoice))
- Sign-up
- email, password, email verification (or Google / GitHub sign-in), company tax ID, phone number and signed MSA before a reserved-cluster bid
- Identity check
- ID on request
- Support
- chat, contact form, email, discord, slack (100+ GPU clusters)
- Enterprise support
- not published
- SLA
- Pass-through only: Terms say Spheron gives no SLA of its own and provider SLAs on dedicated instances go 'up to 99.5%'. Marketing pages claim 99.9% (home, pricing) and 99.99% (on-demand and spot pages); docs say 99.95% for dedicated. Spot has no SLA.
- Status page
- not published
- Published uptime
- not published
Infrastructure
- Regions
27 named regions in 12 countries
Verda Finland 1 / 2 / 3 (Finland) · Spheron ES EU North 1 (Finland) · Spheron ES EU West 1 (France) · Spheron ES UK South 1 (United Kingdom) · Spheron ES ME West 1 (None) · Spheron ES US Central 1 (United States) · Spheron AI Canada 1 / Montreal 1 (Canada) · Spheron AI Norway 1 / Oslo 1 (Norway) · Spheron AI US (Tier 3 DC) (United States) · Spheron MS US (United States) · Spheron MS Calgary (Canada) · Spheron MS Oslo (Norway) · Spheron MS Warsaw (Poland) · Spheron MS Jerusalem (Israel) · Spheron MS Mumbai (India) · Spheron MS Tokyo (Japan) · Spheron MS Singapore (Singapore) · Spheron MS Sydney (Australia) · Spheron AM Michigan 1 (United States) · Massed Compute us-central-1 to -9 · us-east-1 (United States) · Sesterce (North America · EU · India · Japan · Singapore · Australia) (None)- Countries
- United States, Canada, Finland, France, United Kingdom, Norway, Poland, Israel, India, Japan, Singapore, Australia
- Platform
- api, jupyter, ssh, custom_docker_image, persistent_volumes, spot_or_interruptible, bare_metal · SDK : — · Terraform : no · Kubernetes : no · interconnexion : NVLink / NVSwitch on SXM multi-GPU offers · InfiniBand 400 Gb/s NDR on reserved clusters, provider dependent · clusters : Reserved clusters of 8 to 512+ GPUs on monthly terms, by bidding on published clusters (floor price per GPU-hour) or custom request · quote in 24 to 48 h
- Storage
- volume · None · Per-GB-hour volume rates vary by provider and region and are returned only by the authenticated /api/volumes/pricing endpoint and the create-volume wizard; the docs value 0.000137 USD/GB-h for data-crunch is labelled a response example, not a price list.ephemeral · included · Boot/NVMe disk is part of each offer (e.g. 200 to 3300 GB) and included in the hourly rate; Verda dedicated root disk 500 GB or 1 TB.
- Migration out
- Workloads are plain VMs or bare metal with root SSH, so data leaves by rsync or scp from the instance's public IP; no egress fee is published. Volumes are provider-specific (NFS on Verda, virtiofs on Spheron ES, block on Spheron AI, Sesterce and MS) and cannot move between providers or regions; no snapshot or image export is documented. Unused credits are non-refundable below 10,000 USD.
Security & compliance
- SOC 2
- not published
- ISO 27001
- not published
- HIPAA
- not published
- GDPR / DPA
- not published
- Other
- Certifications cited are those of partner data centres: Verda ISO 27001 / GDPR, Sesterce SOC 2 Type II / ISO 27001, Massed Compute HIPAA / SOC 2 Type II. Provider compliance documents shared only under NDA. Operator is a Singapore company (PDPA).
- Legal entity
- Beta360 Pte Ltd
- Funding
- not published
GPUBank market study, read on the provider's own pages · 26 pages read · facts marked "not published" were looked for and not foundPages read ↓
From the study · assessmentSpheron, in short
What stood out when we read the documents: strengths, weaknesses, who it suits, and what we could not find.
Strengths
- Public, unauthenticated offers API (/api/gpu-offers) listing 331 live offers across 24 GPU types and 7 upstream providers at capture
- Low on-demand rates on mid-range cards: RTX A6000 0.5351, RTX 6000 Ada 0.78, L40 0.936, L40S 0.96 USD/GPU-h
- Spot A100 SXM 80GB at 1.1575, H100 SXM at 2.2305 and H200 at 2.9688 USD/GPU-h (Verda capacity)
- Wide geography through resold supply: 12 countries including Israel, India, Japan, Singapore, Australia and Poland
- Stablecoin top-ups (USDC, USDT via CopperX) alongside Stripe cards; credits never expire
- Dedicated public IPv4 per instance and no egress charges
- Reserved cluster bidding against published floor prices, custom clusters of 8 to 512+ GPUs
Weaknesses
- Prices are set by upstream providers and can change on running instances without notice (Terms 5.8 and 5.9)
- 20-minute minimum runtime on every instance, longer provider minimums on some machine types, 60 minutes on 2x MI300X
- Most providers cannot stop an instance: Verda, Sesterce, Spheron MS, Spheron AM and Massed Compute bill the full rate until destroyed
- Credits are non-refundable except claims above 10,000 USD at the company's discretion (Terms 5.2 and 5.3); remaining credits are forfeited on account closure
- No own SLA (pass-through, up to 99.5%) and conflicting uptime figures across pages (99.9%, 99.95%, 99.99%)
- Lowest H100 SXM on-demand rate in the API snapshot was 4.2866 USD/GPU-h and only as 8-GPU nodes; the website 'H100 from 2.98' is an H100 PCIe offer
- No CLI, SDK, Terraform provider or managed Kubernetes documented; API key creation requires whitelisting
- Volume storage prices are not published publicly
- RTX 4090, RTX 5090, GH200 and MI300X had no bookable capacity at capture despite headline rates
Ideal for
- Buyers who want one prepaid account over several second-tier GPU clouds and accept dynamic pricing
- Short, checkpointed jobs on spot A100 / H100 / H200 capacity
- Teams paying in USDC or USDT who accept a Singapore-operated reseller
- Reserved multi-node cluster sourcing (8+ GPUs) through a broker
Not published
- Founding year, funding and headcount of Beta360 Pte Ltd / Spheron: not stated on the official pages read
- Volume storage prices per GB: only via the authenticated API or dashboard
- Reserved cluster floor prices and terms: shown in app.spheron.ai Reserved GPU, not captured
- Terms say reserved is typically 12 months, docs say monthly terms
- Team discount levels (5 to 50% per API docs) are account specific
- Whether identity documents are requested routinely: the Privacy Policy lists government ID and business documents collected 'where required or reasonably necessary'
- the sign-up docs describe no KYC step
- No status page found (status.spheron.ai does not resolve)
- Prices are live marketplace values
- the API snapshot of 2026-09-29 11:39 is one point in time and differed by about a dozen offers from a snapshot a few minutes earlier
- catalogue_count 30 = 20 models in the live API + RTX 4090, RTX 5090, GH200, MI300X (website) + GB200, GB300, R100 (reserve only) + Gaudi 2, H100 NVL, A30 (listed in provider docs, no live offer)
Observations
Per-GPU hourly rates are visible without an account on spheron.network/pricing and /gpu-rental, labelled 'cheapest live on-demand rate' with spot below; they refresh every five minutes and some are marked 'last known'
Home and pricing page showed different specs and rates for the same model minutes apart (H100 on-demand 2.98 on home, spot-only 2.20 on pricing)
The website 'H100' headline 2.98 matches the Massed Compute H100 PCIe offer (20 vCPU, 128 GB RAM, 1250 GB NVMe); SXM5 offers are priced higher
Headline rates for B300 10.37, B200 8.38, RTX PRO 6000 2.35, L40S 0.96, A100 1.43, RTX 6000 Ada 0.78 and the spot rates match the lowest offers in the API snapshot
Pages carry named competitor comparisons (Runpod, Lambda, CoreWeave, AWS, Azure, GCP) with Spheron shown cheaper
Billing is described as per-minute on the marketing site and per-second in the docs; charges settle on a five-minute cycle
Catalogue also lists R100 (ships H2 2026), GB200 and GB300 as reserve-only with custom quotes; home banner advertises 8 x HGX B300 bare-metal nodes in an EU Tier III DC as sold out
Pages read
- spheron.network/ — positioning, headline rates, FAQ, stablecoins, compliance claims
- spheron.network/pricing/ — headline per-GPU rates, minimum runtime, payment FAQ, 99.9% SLA claim
- spheron.network/gpu-rental/ — per-GPU rates incl. last-known RTX 5090 / GH200 / MI300X, RTX 4090
- spheron.network/gpu-rental/h100/ — H100 2.98 headline, SXM5 and PCIe availability
- spheron.network/gpu-rental/mi300x/ — MI300X last-known 3.59, 1x / 2x configs, 60-minute minimum
- spheron.network/gpu-rental/gh200/ — GH200 3.02 last known
- spheron.network/gpu-rental/gb200/ — reserve-only NVL72
- spheron.network/features/on-demand-instances/ — per-minute billing, 99.99% SLA claim
- spheron.network/features/spot-instances/ — up to 50% off
- spheron.network/features/reserved-commitments/ — 8 to 512+ GPUs, 1 to 12+ months
- spheron.network/terms/ — legal entity, pass-through SLA up to 99.5%, non-refundable credits, dynamic pricing, compliance under NDA
- spheron.network/privacy/ — entity, identity verification data categories
- app.spheron.ai/api/gpu-offers?limit=200&page=1 — all per-offer prices, configs, providers, regions (snapshot 11:39
- docs.spheron.ai/billing — credits, per-second, minimum deposit, cards, USDC / USDT via CopperX, stopped billing
- docs.spheron.ai/concepts/regions-providers — providers, GPU tiers, regions, stop support, compliance per provider
- docs.spheron.ai/concepts/instance-types — spot / dedicated / bare metal, 99.95% SLA, minimum runtime
- docs.spheron.ai/concepts/instance-lifecycle — stopped rate, stop support matrix
- docs.spheron.ai/concepts/networking — dedicated public IP, open ports
- docs.spheron.ai/api-reference — endpoints, auth, rate limits, volume pricing endpoint
- docs.spheron.ai/getting-started — sign-up methods, deposit limits
- docs.spheron.ai/reserved-gpus — bidding, bid hold, MSA, tax ID, phone
- docs.spheron.ai/overview — no egress fees, provider compliance
- docs.spheron.ai/general-info — support channels
- docs.spheron.ai/concepts/teams — API key whitelist
- docs.spheron.ai/templates — 7 startup scripts
- docs.spheron.ai/changelog — Kubernetes add-on removed with Voltage Park
Questions about Spheron
What GPUs does Spheron offer?
Spheron has 23 GPU types with a fresh offer today, for example H100 from $2.21, A100 80GB from $1.19, H200 from $4.80, RTX PRO 6000 from $2.35, L40S from $1.80, B200 from $11.12 per GPU-hour on-demand. The full list with prices and market medians is on this page.
How does Spheron bill?
Billing granularity: mixed. Minimum charge: 20-minute minimum runtime on every instance (60 minutes on 2x MI300X); some machine types carry a longer provider minimum shown in the deploy wizard, charged in full on early termination Stopped instance: Stop exists only on Spheron AI (some machine types and regions) and Spheron ES; the stopped rate covers retained disk plus public IP where the provider charges for it (API example: 0.0412 USD/h for a 512 GB root disk and IP). On Verda, Sesterce, Spheron MS, Spheron AM, Massed Compute, all bare metal and all spot, an instance bills the full hourly rate until destroyed.
How trustworthy is Spheron?
Its Trust Score is 53/100 (v0.1), read from its own documents. Strong on pricing & billing transparency. Weaker on company & track record, infrastructure control, reliability signals, security & compliance. The score measures transparency and published guarantees, not customer satisfaction; every criterion and its source is listed above.
Where are Spheron's data centres?
27 named regions in 12 countries: United States, Canada, Finland, France, United Kingdom, Norway, Poland, Israel, India, Japan, Singapore, Australia.
How do you pay Spheron, and are there credits?
Payment by card, crypto. No sign-up credit is published. Identity check: ID on request.