Home / Compute
ParalonCloud
paraloncloud.com/pricing/rtx-4090GPU marketplacecompute/paraloncloudJSONcomparereport an errorprovider JSON
- Free allowance
- No free GPU tier is published; the RTX 4090 page read on 2026-11-03 gives a price per hour of USD 0.41 to USD 0.45, set by independent providers, billed by the minute with no contract or minimum spend; the rent page lists H100, A100 and L40S models at list prices (H100 SXM USD 2.40 per hour...
- Verdict
- paidno free allowance; listed for its price
- Cost
- no price stated
- Identity
- emailno card, no phone
- Binding limit
- noneno named constraint binds
- Evidence
- documented11 sources; a first-party page was read and the sentence quoted
- Freshness
- stale on 2026-12-04checked 2026-11-03 (live); stale on 2026-12-04
| Free tier | no free GPU tier is published; the RTX 4090 page read on 2026-11-03 gives a price per hour of USD 0.41 to USD 0.45, set by independent providers, billed by the minute with no contract or minimum spend; the rent page lists H100, A100 and L40S models at list prices (H100 SXM USD 2.40 per hour, A100 80GB PCIe USD 1.90, L40S USD 0.65) with none available right now, and the pricing page's GPU table is empty in a static read. The same account has a free trial of 250,000 inference tokens on a free key. unit none; metered on wall-clock; the GPU table on the pricing page needs JavaScript; the card page was read instead. Spot machines can be taken back by the owner with two minutes' notice |
|---|
| Price | none stated |
|---|
| Binding | none No quota is stated; a 1-hour balance is needed up front and a rental stops when credits cannot cover the next minute. |
|---|
| Identity | email Sign in with Google or a one-time email code; a short 1-hour balance is required up front, topped up by card, by a USDC transfer or from a connected USDC or USDT wallet. No phone or ID document is stated on the pages read. |
|---|
| Egress | not published on the pages read |
|---|
| Interface | ssh inferred from the docs, which describe Jupyter Lab in the browser or SSH from a terminal; not probed |
|---|
| Tooling | JupyterLab or SSH on a rented machine; an OpenAI-compatible inference API with prlc_ keys; a Blender rendering beta |
|---|
| Composes with | Per-minute RTX 4090 rentals from independent hosts plus an inference API on the same credit balance. |
|---|
| Reliability | The docs state that a rental not answering within 5 minutes is closed without charge, and that after 15 minutes of failure it is ended; a machine whose credits run out is stopped and wiped. |
|---|
| Notes | Found via computeprices.com (provider slug paraloncloud). The aggregator lists RTX 4090 at USD 0.420 (1x) and USD 0.450 (2x) plus RTX 4060 Ti, 4070 SUPER, 4080 SUPER, 5060 Ti, 5070 Ti and 5090. The five cards asked about are not listed except the RTX 4090. |
|---|
| Parsed numbers | |
|---|
Evidence 11
Unknown values 3
- egress metered: egress metering was not published
- per object cap stated: the field does not apply to this kind of offer
- retention stated: the field does not apply to this kind of offer