# Vast.ai

Official pricing: https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx  
Category: gpu-cloud · Isolation: container

## Pricing regimes (raw)

- **RTX4090, on-demand** (resource) [alt]
- **RTX5090, on-demand** (resource) [alt]
- **RTX-PRO-6000, on-demand** (resource) [alt]
- **A100-80GB, on-demand** (resource) [alt]
- **H100, on-demand** (resource) [alt]
- **H200, on-demand** (resource) [alt]
- **B200, on-demand** (resource) [alt]
- **RTX4090, spot** (resource) [alt, spot]

## Features

Yes: snapshots, pause/resume, persistent disk, volumes, ≥24 h sessions, custom image (Docker or snapshot), start from your own snapshot, your own Docker/OCI image, browser, desktop GUI, residential IPs, SSH, HTTPS preview URLs, open internet, static egress IP, SOC 2, HIPAA, GPU, arm64, Python, EU data residency, snapshots on demand, inbound access rules, firewall inside (nftables), extra volumes

No: memory snapshots, fork/clone, idle auto-stop, full VM (own kernel), Docker inside, nested virtualization, computer-use API, browser + desktop control, anti-bot stealth, CAPTCHA solving, preinstalled agents, public IPv4, egress allowlist, self-hosting, BYOC, open source, SSO, Windows, macOS, Node.js, HTTP method/path egress rules, wake on request, live resize, live fork (no pause), memory fork, MCP server, automatic snapshots, agent harness API, hosted agent API (their own agent), gVisor or VM (no shared kernel), secret proxy, secret proxy for any API, shared volumes

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

## Caveats

- No universal list rate: GPU, CPU/RAM share, bandwidth and storage are host-offer-specific. Per-second compute; storage continues while stopped. Serverless uses underlying marketplace workers with no extra surcharge. Reserved prepayment up to 50% off is not a fixed tariff. Dated public API snapshots added for 1 GPU, verified hosts, reliability>=0.98; first two cheapest offers by total listed hourly price. dph_base is compute, dph_total adds default 10GB storage; bandwidth remains extra. gpu_frac describes host share, NOT fractional physical GPU.
- 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.
- 2026-09-28 fix pass: snapshot regimes collapsed to one per GPU (cheapest live verified offer, flag stock). Disk priced at $0.2667/GB-month = median host storage_cost across 71 live verified 1-GPU offers (range $0.08-$1.00/GB-month; the host you rent sets it; storage keeps billing while stopped). Egress $0.0052/GB = median host inet_up_cost across the same offers (host-specific; download also billed per host). Live prices moved since the earlier snapshot (e.g. H100 SXM $1.87 -> $3.09/h, H200 $3.55 -> $4.80/h).

## How this provider charges

As of 2026-09-28. Marketplace on-demand/reserved/interruptible and serverless containers.

No universal list rate: GPU, CPU/RAM share, bandwidth and storage are host-offer-specific. Per-second compute; storage continues while stopped. Serverless uses underlying marketplace workers with no extra surcharge. Reserved prepayment up to 50% off is not a fixed tariff. Dated public API snapshots added for 1 GPU, verified hosts, reliability>=0.98; first two cheapest offers by total listed hourly price. dph_base is compute, dph_total adds default 10GB storage; bandwidth remains extra. gpu_frac describes host share, NOT fractional physical GPU.

| Regime | When applicable | Billing | Numbers ($/GPU-hour unless slice) | Source |
|---|---|---|---|---|
| on-demand  | RTX4090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| on-demand  | RTX5090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| on-demand  | RTX-PRO-6000 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| on-demand  | A100-80GB ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| on-demand  | H100 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| on-demand  | H200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| on-demand  | B200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| spot  | RTX4090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| spot  | RTX5090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| spot  | RTX-PRO-6000 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| spot  | A100-80GB ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| spot  | H100 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| spot  | H200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| spot  | B200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| reserved  | RTX4090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| reserved  | RTX5090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| reserved  | RTX-PRO-6000 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| reserved  | A100-80GB ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| reserved  | H100 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| reserved  | H200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| reserved  | B200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| serverless  | RTX4090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| serverless  | RTX5090 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| serverless  | RTX-PRO-6000 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| serverless  | A100-80GB ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| serverless  | H100 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| serverless  | H200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| serverless  | B200 ; count None; Region/availability must be checked at launch. | per-second; min unknowns | unknown; dynamic-marketplace; per_gpu_hour | [source](https://github.com/vast-ai/docs/blob/main/guides/pricing.mdx) |
| on-demand  | H200 H200; count [1]; Georgia, US | per-second; min unknowns | 3.5526315789473686; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22H200%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | H200 H200; count [1]; New Jersey, US | per-second; min unknowns | 3.5526315789473686; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22H200%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | RTX-PRO-6000 RTX PRO 6000 S; count [1]; California, US | per-second; min unknowns | 1.4000000000000001; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22RTX+PRO+6000+S%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | RTX-PRO-6000 RTX PRO 6000 S; count [1]; Australia, AU | per-second; min unknowns | 1.4400000000000002; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22RTX+PRO+6000+S%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | RTX5090 RTX 5090; count [1]; United States, US | per-second; min unknowns | 0.44; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22RTX+5090%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | RTX5090 RTX 5090; count [1]; South Korea, KR | per-second; min unknowns | 0.4666666666666666; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22RTX+5090%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | B200 B200; count [1]; Virginia, US | per-second; min unknowns | 5.76875; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22B200%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | B200 B200; count [1]; Oregon, US | per-second; min unknowns | 6.25; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22B200%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | A100-40GB A100 SXM4; count [1]; California, US | per-second; min unknowns | 0.4266666666666667; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22A100+SXM4%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | A100-40GB A100 SXM4; count [1]; Alberta, CA | per-second; min unknowns | 0.5866666666666667; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22A100+SXM4%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | RTX4090 RTX 4090; count [1]; Michigan, US | per-second; min unknowns | 0.38666666666666666; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22RTX+4090%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | RTX4090 RTX 4090; count [1]; California, US | per-second; min unknowns | 0.39999999999999997; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22RTX+4090%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | H100 H100 SXM; count [1]; Germany, DE | per-second; min unknowns | 1.8666666666666665; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22H100+SXM%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |
| on-demand  | H100 H100 SXM; count [1]; Germany, DE | per-second; min unknowns | 1.8666666666666665; dynamic-snapshot; per_gpu_hour | [source](https://console.vast.ai/api/v0/bundles/?q=%7B%22verified%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rentable%22%3A+%7B%22eq%22%3A+true%7D%2C+%22rented%22%3A+%7B%22eq%22%3A+false%7D%2C+%22reliability%22%3A+%7B%22gte%22%3A+0.98%7D%2C+%22num_gpus%22%3A+%7B%22eq%22%3A+1%7D%2C+%22gpu_name%22%3A+%7B%22eq%22%3A+%22H100+SXM%22%7D%2C+%22type%22%3A+%22on-demand%22%2C+%22order%22%3A+%5B%5B%22dph_total%22%2C+%22asc%22%5D%5D%2C+%22limit%22%3A+2%7D) |

## Gotchas
- No universal list rate: GPU, CPU/RAM share, bandwidth and storage are host-offer-specific. Per-second compute; storage continues while stopped. Serverless uses underlying marketplace workers with no extra surcharge. Reserved prepayment up to 50% off is not a fixed tariff. Dated public API snapshots added for 1 GPU, verified hosts, reliability>=0.98; first two cheapest offers by total listed hourly price. dph_base is compute, dph_total adds default 10GB storage; bandwidth remains extra. gpu_frac describes host share, NOT fractional physical GPU.
- 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.