# Fluidstack

Official pricing: https://fluidstack.io/  
Category: gpu-cloud · Isolation: dedicated-host

## Pricing regimes (raw)

- **H100, reserved** (resource) [alt, sales]
- **H200, reserved** (resource) [alt, sales]
- **B200, reserved** (resource) [alt, sales]
- **GB200, reserved** (resource) [alt, sales]

## Features

Yes: pause/resume, persistent disk, ≥24 h sessions, custom image (Docker or snapshot), full VM (own kernel), Docker inside, nested virtualization, SSH, SSO, GPU, Python, Node.js, gVisor or VM (no shared kernel), firewall inside (nftables)

No: snapshots, memory snapshots, fork/clone, volumes, idle auto-stop, start from your own snapshot, your own Docker/OCI image, browser, desktop GUI, computer-use API, browser + desktop control, anti-bot stealth, CAPTCHA solving, residential IPs, preinstalled agents, public IPv4, HTTPS preview URLs, egress allowlist, static egress IP, self-hosting, BYOC, open source, arm64, Windows, macOS, 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), inbound access rules, secret proxy, secret proxy for any API, extra volumes, shared volumes

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

## Caveats

- No public offering at all as of 2026-09-29: fluidstack.io is now a mission/recruiting site (no product, pricing, sign-up or contact-sales link); /pricing and /gpu-cloud return 404 (Wayback: /pricing 404 since at least 2026-01-28). The platform (docs.fluidstack.io: managed Kubernetes/Slurm, alpha instances API) serves existing contracted customers via a Technical Account Manager. Last published list prices (Wayback 2025-10-17, historic, not current): H100 SXM $2.10, H200 SXM $2.30, A100 80GB $1.30 per GPU-hour; B200/GB200 'On Request'.
- 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.
- Modes with gpu=null are intentionally excluded, not free GPU modes. Never substitute 0 for a null rate. Flags alt indicate DIY/ML infrastructure, not the same thing as managed untrusted-code sandboxes.
- Unknown fee, storage or bandwidth means full monthly total may be unavailable. Rates exclude taxes.
- Instances API (v1alpha1, regional https://<region>.atlas.fluidstack.io) supports preemptible and ephemeral instances, CPU instance types (CLI default 'cpu.2x') and GPU types like 'h100.8x', plus start/stop. Prices, billing increment and preemptible discount are not published; access is contract-only.
- Contract terms (T&Cs): fees set per Order, 'payable in advance', 'non-cancellable, non-refundable' unless the Order says otherwise; per-minute/per-hour fees may require an advance deposit and access 'may cease' if funds run out (hard stop); subscriptions auto-renew; renewal price increases of at least 3%/year allowed with 30 days notice.

## How this provider charges

As of 2026-09-28. Dedicated GPU infrastructure / clusters.

Current site is enterprise AI infrastructure, no public self-serve GPU tariff. Historic $1.99 H100 claims not accepted as current. Sandbox-like on-demand product availability not verified.

| Regime | When applicable | Billing | Numbers ($/GPU-hour unless slice) | Source |
|---|---|---|---|---|
| reserved  | H100 ; count None; Region/availability must be checked at launch. | unknown; min unknowns | unknown; sales-unpublished; per_gpu_hour | [source](https://fluidstack.io/) |
| reserved  | H200 ; count None; Region/availability must be checked at launch. | unknown; min unknowns | unknown; sales-unpublished; per_gpu_hour | [source](https://fluidstack.io/) |
| reserved  | B200 ; count None; Region/availability must be checked at launch. | unknown; min unknowns | unknown; sales-unpublished; per_gpu_hour | [source](https://fluidstack.io/) |
| reserved  | GB200 ; count None; Region/availability must be checked at launch. | unknown; min unknowns | unknown; sales-unpublished; per_gpu_hour | [source](https://fluidstack.io/) |

## Gotchas
- Current site is enterprise AI infrastructure, no public self-serve GPU tariff. Historic $1.99 H100 claims not accepted as current. Sandbox-like on-demand product availability not verified.
- 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.