# Hyperbolic

Official pricing: https://www.hyperbolic.ai/marketplace  
Category: gpu-cloud · Isolation: vm

## Pricing regimes (raw)

- **H100 SXM · on-demand (marketplace floor)** (resource): $0/vCPU-h, $0/GiB-h
- **H200 · on-demand (marketplace floor)** (resource): $0/vCPU-h, $0/GiB-h
- **B200 · on-demand (marketplace floor)** (resource): $0/vCPU-h, $0/GiB-h
- **Reserved term (prepaid, 7/14/30 days)** (resource): $0/vCPU-h, $0/GiB-h
- **Private Cloud (single-tenant, >1 year contract)** (resource) [sales]

## Features

Yes: persistent disk, volumes, ≥24 h sessions, full VM (own kernel), Docker inside, root, SSH, public IPv4, raw TCP inbound, GPU, Python, EU data residency, gVisor or VM (no shared kernel), inbound access rules, firewall inside (nftables), extra volumes, shared volumes

No: snapshots, memory snapshots, fork/clone, pause/resume, idle auto-stop, start from your own snapshot, your own Docker/OCI image, nested virtualization, browser, desktop GUI, computer-use API, browser + desktop control, anti-bot stealth, CAPTCHA solving, residential IPs, preinstalled agents, HTTPS preview URLs, egress allowlist, static egress IP, self-hosting, BYOC, open source, SOC 2, HIPAA, SSO, arm64, Windows, macOS, Node.js, HTTP method/path egress rules, wake on request, live resize, live fork (no pause), memory fork, MCP server, automatic snapshots, snapshots on demand, agent harness API, hosted agent API (their own agent), secret proxy, secret proxy for any API

Unknown: everything else. Evidence (source + quote) per feature: https://battleships.dev/data/providers/hyperbolic.json → feature_evidence

## Caveats

- Old hyperbolic.xyz/pricing returns 404. New official marketplace has weekly refreshed starting prices. No charges for failed instances. Reservations paid upfront; <1 year reserved and >1 year Private Cloud.
- Any null count, CPU/RAM, billing increment or minimum is unverified, not unlimited/free.
- This is GPU-product research, not a claim that every feature of the entire vendor documentation was audited. Unverified features are null.
- GPU multi-count bundles, variant/region selection and fractional slices require a SKU-aware estimator; unsupported rows remain unpriced in strict cards.
- On-demand GPU rates are marketplace floor ("starting at") prices, refreshed weekly from supplier bids: actual availability and price at launch may be higher. Host CPU/RAM is bundled and unspecified.
- Unknown fee, storage or bandwidth means full monthly total may be unavailable. Rates exclude taxes.
- Reservation Protection is ON by default: at the end of a reserved term the instance keeps running and bills at the current on-demand rate until terminated.
- On-demand price is locked per instance at creation, but Hyperbolic reserves the right to contact long-running below-market instances about a price increase.
- Onboard local NVMe is included free with every instance (size varies by configuration; up to 24 TB per 8-GPU H100/H200 node) and is erased on termination.

## How this provider charges

As of 2026-09-28. GPU marketplace VMs/bare metal and reserved clusters.

Old hyperbolic.xyz/pricing returns 404. New official marketplace has weekly refreshed starting prices. No charges for failed instances. Reservations paid upfront; <1 year reserved and >1 year Private Cloud.

| Regime | When applicable | Billing | Numbers ($/GPU-hour unless slice) | Source |
|---|---|---|---|---|
| on-demand  | H100 SXM; count None; Region/availability must be checked at launch. | unknown; min unknowns | 3.19; list-from; per_gpu_hour | [source](https://www.hyperbolic.ai/marketplace) |
| on-demand  | H200 ; count None; Region/availability must be checked at launch. | unknown; min unknowns | 3.99; list-from; per_gpu_hour | [source](https://www.hyperbolic.ai/marketplace) |
| on-demand  | B200 ; count None; Region/availability must be checked at launch. | unknown; min unknowns | 5.99; list-from; per_gpu_hour | [source](https://www.hyperbolic.ai/marketplace) |

## Gotchas
- Old hyperbolic.xyz/pricing returns 404. New official marketplace has weekly refreshed starting prices. No charges for failed instances. Reservations paid upfront; <1 year reserved and >1 year Private Cloud.
- Any null count, CPU/RAM, billing increment or minimum is unverified, not unlimited/free.
- This is GPU-product research, not a claim that every feature of the entire vendor documentation was audited. Unverified features are null.
- GPU multi-count bundles, variant/region selection and fractional slices require a SKU-aware estimator; unsupported rows remain unpriced in strict cards.

## Worked example (required protocol workload)
4 vCPU / 8 GiB, 50 concurrent × 8 hours/day × 22 days = **8,800 instance-hours/month**. 30% CPU utilization does not reduce allocated GPU uptime. 50 GiB snapshots and 100 GiB egress are additional. The requested CPU-only workload has no GPU type/count, so its full GPU-provider total is **null**, not a fictitious CPU equivalent.

No fixed single-full-GPU USD quote suitable for a numeric example was verified. Compute = 8,800 × selected offer $/hour, or required node count × node price; keep the estimate null until the offer/FX/commitment is resolved.

## Evidence scope
Sources read are listed in the provider/features JSON. Raw pricing pages, relevant docs and API payloads are retained under `../raw/`. All unknown feature toggles remain null.