fshy

fshy / Compute

Prime Intellect

primeintellect.ai · Compute · 2 records · JSON

Prime Intellect Sandboxes Compute

paidemail?card?Drequest-rate

www.primeintellect.ai/sandboxes · container and VM sandbox · compute/prime-intellect · JSON · compare

Free tierno free tier: tier-free per-second pricing at $0.02 a vCPU-hour, $0.0125 a GiB-hour of memory and $0.0002 a GiB-hour of disk, with no subscription or minimum spend
unit none; metered on none
Pricenone stated
Per-object capnot stated
Retentionnot stated
Egressnot stated
Identityemail (placeholder) not published
Limits
  • request-rate $0.02 a vCPU-hour, $0.0125 a GiB-hour of memory and $0.0002 a GiB-hour of disk source
  • request-rate 100 concurrent sandboxes in the worked example; the page advertises scaling to tens of thousands source
not stated
Bindingrequest-rate
Interfaceunknown
Toolingnot stated
Composes withnot stated
Reliabilityno reliability event recorded
Evidence rungdocumented the sandboxes page was read on 2026-10-22 and publishes the three meters and a worked rate.
Checked2026-10-23 (live); stale on 2026-11-22
Evidence (2)
datesourcewhat it proved
2026-10-02www.primeintellect.ai/sandboxesnot published
2026-10-22www.primeintellect.ai/sandboxes'Tier-free pricing built for scale, with no subscriptions or minimum spend' with 'vCPU $0.02 / hour', 'Memory $0.0125 / GiB / hour', 'Disk $0.0002 / GiB / hour', and '1 vCPU · 1 GiB RAM · 5 GiB disk · 100 Sandboxes $3.35 / hour = $0.0335 / sandbox / hour', quoted
More fields (15)
{
 "talaria_m10h_usd": 27.0,
 "talaria_m24h_usd": 64.8,
 "nemo_m10h_usd": 28.2,
 "keep_rate": 1.0,
 "keep_rate_basis": null,
 "billing_model": "unclassified",
 "shape": {
  "vcpu": 1,
  "ram": 1,
  "disk": 5,
  "gpu_class": null,
  "gpu_count": null,
  "note": "the worked example: 1 vCPU, 1 GiB and 5 GiB of disk at $0.0335 a sandbox-hour"
 },
 "session_cap": {
  "wall_clock_min": null,
  "idle_reap_s": null,
  "restart_required": false,
  "note": "every workload stays in an isolated microVM; the page states no reap"
 },
 "persistent_storage": null,
 "network_out": null,
 "price_note": null,
 "free_tier_kind": "none",
 "burst_x": 1.0,
 "ps": {
  "gib": 5,
  "price": "$0.0002 a GiB-hour",
  "durability": "sandbox disk",
  "on_stop": true,
  "on_destroy": false,
  "retention": "until deleted"
 },
 "net": {
  "out": true,
  "in": false,
  "ports": null,
  "static_ip": null,
  "ipv6": null,
  "allowance": null
 }
}
Unknown or withheld values (17)
  • $.chaining not-stated
  • $.egress.metered egress-unknown
  • $.egress.terms not-stated
  • $.evidence[0].proved process
  • $.extras.keep_rate_basis process
  • $.extras.price_note process
  • $.identity_cost.detail defaulted
  • $.identity_cost.kind identity-defaulted
  • $.interface interface-unknown
  • $.interface_detail withheld
  • $.limits_note not-stated
  • $.per_object_cap.notes process
  • $.per_object_cap.stated not-stated
  • $.reliability_note none-observed
  • $.retention.notes not-extracted
  • $.retention.stated not-stated
  • $.tooling not-stated

Prime Intellect Compute

deademailno cardDvolume-quota

www.primeintellect.ai/blog/fast-compute-grants · GPU cloud · compute/prime-intellect-2 · JSON · compare

Free tierThe grant program is gone. Its own page now returns HTTP 404, and no signup credit exists: the docs describe a prepaid model where 'your Prime Intellect credits are deducted every minute while the Pod is active' and 'If your credits run out, your Pods will be automatically deleted'.
unit none; metered on wall-clock; kind not stated
Pricenone stated
Per-object capnot applicable
Retentionnot applicable
Egressnot stated
Identityemail not published
Limits
  • volume-quota Both doors fail: there is no signup credit, so the first pod needs a funded balance, and the USD 500 to USD 100,000 Fast Compute Grants program no longer has a live page. source
Both doors fail: there is no signup credit, so the first pod needs a funded balance, and the USD 500 to USD 100,000 Fast Compute Grants program no longer has a live page.
Bindingvolume-quota Both doors fail: there is no signup credit, so the first pod needs a funded balance, and the USD 500 to USD 100,000 Fast Compute Grants program no longer has a live page.
Interfaceunknown inferred from the tooling the provider documents; not probed
Toolingnot stated
Composes withNone while the grant is closed. If it reopened it would be a real credit lane, because the platform aggregates other providers and accepts prepaid balances.
Reliabilityno reliability event recorded
NotesREFUTATION of a widely recirculated grant. Two figures circulate: 'USD 500 to USD 100,000 Fast Compute Grants' and 'USD 50 free compute at ICLR'. Both source pages return 404 as of 2026-10-22, and the ICLR page carried an expiry of 31 December 2024. The grant was real in 2024; it is not an offer now. What would settle it: a live page on primeintellect.ai naming an amount and an application route.
Evidence rungdocumented
Checked2026-10-22 (live); stale on 2026-11-21
Evidence (4)
datesourcewhat it proved
2026-10-22www.primeintellect.ai/blog/fast-compute-grantsfetch --head returned 'HTTP/1.1 404 Not Found'; the page previously offered 'USD 500-USD 100k worth of Prime Intellect compute credits'
2026-10-22www.primeintellect.ai/grantsfetch --head returned 'HTTP/1.1 404 Not Found'
2026-10-22docs.primeintellect.ai/faqquoted: 'Your Prime Intellect credits are deducted every minute while the Pod is active. If your credits run out, your Pods will be automatically deleted'; no free tier described
2026-10-22buildaicurrent.com/credits/prime-intellect-free-credits/independent check dated 2026-08-21 records that both the grants and ICLR pages returned 404 and that no live official page states either offer
More fields (1)
{
 "wave2": {
  "class": "gpu-marketplace-grant",
  "free_tier_kind": null,
  "shape": {
   "vcpu": null,
   "ram_gib": null,
   "gpu_class": null,
   "gpu_count": null
  }
 }
}
Unknown or withheld values (10)
  • $.egress.metered egress-unknown
  • $.egress.terms not-stated
  • $.identity_cost.detail process
  • $.interface interface-unknown
  • $.per_object_cap.notes process
  • $.per_object_cap.stated not-applicable
  • $.reliability_note none-observed
  • $.retention.notes process
  • $.retention.stated not-applicable
  • $.tooling not-stated