Runcrate
- Free allowance
- No free GPU tier is published; the live rate card read on 2026-11-03 lists 19 GPU types from USD 0.36 an hour: H100 80GB USD 3.86, A100 80GB USD 1.73, L40S 48GB USD 0.54, L40 USD 0.47, RTX 4090 USD 0.57, H200 USD 4.63, B200 USD 7.67.
- Verdict
- paidno free allowance; listed for its price
- Cost
- no price stated
- Identity
- cardcard required, no phone
- Binding limit
- write-gateda launch needs USD 10 or an hour of the machine, whichever is more; a balance is added by card from USD 20
- Evidence
- documented7 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 live rate card read on 2026-11-03 lists 19 GPU types from USD 0.36 an hour: H100 80GB USD 3.86, A100 80GB USD 1.73, L40S 48GB USD 0.54, L40 USD 0.47, RTX 4090 USD 0.57, H200 USD 4.63, B200 USD 7.67. No L4 is listed. Billed by the second on most GPUs and by the minute on a few; booting is never billed. unit none; metered on wall-clock; prepaid terms of 1 to 12 months are up to 15% off; the aggregator shows H100 SXM at USD 2.49 where the provider rate card shows USD 3.86 |
|---|---|
| Price | none stated |
| Limits |
A launch is gated on a prepaid balance added by card; there is no free allowance. |
| Identity | card The sign-up form asks for name, email, password and an age or consent checkbox; balance is added by card before use, from USD 20 to USD 10,000 at a time. Reserved multi-node clusters need a verification of the company and the need. No phone or ID document is stated for self-serve. |
| Egress | not published on the pages read |
| Interface | rest, cli, ssh taken from the pricing page, which lists SSH, a CLI and an unauthenticated /api/gpu-pricing endpoint; not probed |
| Tooling | SSH, Jupyter and VS Code in the browser, Python and TypeScript SDKs, a CLI and an MCP server; an OpenAI-compatible inference API on the same balance |
| Composes with | One prepaid balance pays for GPU hours, token-billed inference and storage volumes; the public /api/gpu-pricing endpoint publishes the live rate card without a key. |
| Reliability | Capacity is sourced multi-cloud rather than from owned data centres per the aggregator profile; sold-out GPUs stay listed for 30 days at 0 available per the rate card. |
| Notes | Found via computeprices.com (provider slug runcrate). The aggregator row for L4 (USD 1.24) is not on the provider rate card read 2026-11-03. Not read: the L4 availability. |
| Parsed numbers |
|
Evidence 7
| date | source | what it proved |
|---|---|---|
| 2026-11-03 | www.runcrate.ai/pricing | H100 80GB rate is USD 3.86 an hour |
| 2026-11-03 | www.runcrate.ai/pricing | A100 80GB rate is USD 1.73 an hour |
| 2026-11-03 | www.runcrate.ai/pricing | L40S rate is USD 0.54 an hour |
| 2026-11-03 | www.runcrate.ai/pricing | RTX 4090 rate is USD 0.57 an hour |
| 2026-11-03 | www.runcrate.ai/pricing | balance is added by card from USD 20; a launch needs USD 10 |
| 2026-11-03 | www.runcrate.ai/auth/signup | sign-up asks for name, email and password |
| 2026-11-03 | computeprices.com/providers/runcrate | the aggregator lists an L4 for Runcrate at USD 1.24 an hour |
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