# GPU Compare — Full Documentation

> This is the detailed companion to [llms.txt](https://gpucompare.com/llms.txt). It contains complete pricing tables, GPU specifications, provider statistics, and URL slug mappings for all resources on GPU Compare.

## Overview

GPU Compare aggregates real-time pricing from 14 cloud GPU providers. All pricing data is refreshed every 5 minutes.

- **Total active offers**: 325+
- **Providers with active offers**: 14
- **GPU models covered**: 17
- **Price range**: $N/A/hr – $8.18/hr
- **Currency**: USD (hourly rates)

## URL Structure & Slug Format

All GPU page URLs follow this pattern:
```
https://gpucompare.com/gpu/{slug}
```

Slugs are the GPU name in lowercase with spaces replaced by hyphens. Examples:
- "H100 SXM5" → `h100-sxm5` → https://gpucompare.com/gpu/h100-sxm5
- "RTX 4090" → `rtx-4090` → https://gpucompare.com/gpu/rtx-4090
- "A100 80GB SXM" → `a100-80gb-sxm` → https://gpucompare.com/gpu/a100-80gb-sxm

Provider page URLs:
```
https://gpucompare.com/providers/{provider-slug}
```

### Markdown Endpoints

Append `.md` to any GPU or provider URL to get a clean, LLM-friendly markdown version with structured data, specs, and pricing tables — no HTML, no JavaScript, no navigation chrome.

```
https://gpucompare.com/gpu/{slug}.md          → GPU specs + pricing table in markdown
https://gpucompare.com/providers/{slug}.md    → Provider overview + offers in markdown
```

Examples:
- https://gpucompare.com/gpu/nvidia-h100-sxm5.md
- https://gpucompare.com/providers/runpod.md

## Complete GPU Slug Reference

- `nvidia-b200` → [NVIDIA B200](https://gpucompare.com/gpu/nvidia-b200)
- `nvidia-h200-sxm5` → [NVIDIA H200 SXM5](https://gpucompare.com/gpu/nvidia-h200-sxm5)
- `nvidia-a100-80gb-sxm4` → [NVIDIA A100 80GB SXM4](https://gpucompare.com/gpu/nvidia-a100-80gb-sxm4)
- `nvidia-h100-sxm5` → [NVIDIA H100 SXM5](https://gpucompare.com/gpu/nvidia-h100-sxm5)
- `nvidia-rtx-6000-ada` → [NVIDIA RTX 6000 Ada](https://gpucompare.com/gpu/nvidia-rtx-6000-ada)
- `nvidia-l40` → [NVIDIA L40](https://gpucompare.com/gpu/nvidia-l40)
- `nvidia-rtx-a6000` → [NVIDIA RTX A6000](https://gpucompare.com/gpu/nvidia-rtx-a6000)
- `nvidia-a40` → [NVIDIA A40](https://gpucompare.com/gpu/nvidia-a40)
- `nvidia-l40s` → [NVIDIA L40S](https://gpucompare.com/gpu/nvidia-l40s)
- `nvidia-rtx-5090` → [NVIDIA RTX 5090](https://gpucompare.com/gpu/nvidia-rtx-5090)
- `nvidia-v100-32gb-sxm2` → [NVIDIA V100 32GB SXM2](https://gpucompare.com/gpu/nvidia-v100-32gb-sxm2)
- `nvidia-l4` → [NVIDIA L4](https://gpucompare.com/gpu/nvidia-l4)
- `nvidia-a10` → [NVIDIA A10](https://gpucompare.com/gpu/nvidia-a10)
- `nvidia-rtx-4090` → [NVIDIA RTX 4090](https://gpucompare.com/gpu/nvidia-rtx-4090)
- `nvidia-rtx-a5000` → [NVIDIA RTX A5000](https://gpucompare.com/gpu/nvidia-rtx-a5000)
- `nvidia-a16` → [NVIDIA A16](https://gpucompare.com/gpu/nvidia-a16)
- `nvidia-rtx-a4000` → [NVIDIA RTX A4000](https://gpucompare.com/gpu/nvidia-rtx-a4000)

## GPU Pricing Table

| GPU | Min Price | Max Price | Providers | Offers |
|-----|-----------|-----------|-----------|--------|
| [NVIDIA B200](https://gpucompare.com/gpu/nvidia-b200) | $6.52/hr | $48.00/hr | 4 | 6 |
| [NVIDIA H200 SXM5](https://gpucompare.com/gpu/nvidia-h200-sxm5) | $3.44/hr | $43.16/hr | 6 | 11 |
| [NVIDIA A100 80GB SXM4](https://gpucompare.com/gpu/nvidia-a100-80gb-sxm4) | $0.73/hr | $26.40/hr | 8 | 57 |
| [NVIDIA H100 SXM5](https://gpucompare.com/gpu/nvidia-h100-sxm5) | $1.92/hr | $30.80/hr | 9 | 40 |
| [NVIDIA RTX 6000 Ada](https://gpucompare.com/gpu/nvidia-rtx-6000-ada) | $0.79/hr | $65.44/hr | 5 | 27 |
| [NVIDIA L40](https://gpucompare.com/gpu/nvidia-l40) | $0.88/hr | $8.00/hr | 3 | 13 |
| [NVIDIA RTX A6000](https://gpucompare.com/gpu/nvidia-rtx-a6000) | $0.43/hr | $8.49/hr | 5 | 18 |
| [NVIDIA A40](https://gpucompare.com/gpu/nvidia-a40) | $0.57/hr | $0.60/hr | 1 | 2 |
| [NVIDIA L40S](https://gpucompare.com/gpu/nvidia-l40s) | $0.60/hr | $12.65/hr | 7 | 36 |
| [NVIDIA RTX 5090](https://gpucompare.com/gpu/nvidia-rtx-5090) | $0.10/hr | $6.16/hr | 5 | 13 |
| [NVIDIA V100 32GB SXM2](https://gpucompare.com/gpu/nvidia-v100-32gb-sxm2) | $9.36/hr | $10.30/hr | 2 | 2 |
| [NVIDIA L4](https://gpucompare.com/gpu/nvidia-l4) | $1.90/hr | $8.36/hr | 3 | 9 |
| [NVIDIA A10](https://gpucompare.com/gpu/nvidia-a10) | $0.48/hr | $1.42/hr | 4 | 7 |
| [NVIDIA RTX 4090](https://gpucompare.com/gpu/nvidia-rtx-4090) | $0.16/hr | $4.31/hr | 5 | 17 |
| [NVIDIA RTX A5000](https://gpucompare.com/gpu/nvidia-rtx-a5000) | $0.09/hr | $6.25/hr | 4 | 4 |
| [NVIDIA A16](https://gpucompare.com/gpu/nvidia-a16) | $0.47/hr | $4.50/hr | 3 | 42 |
| [NVIDIA RTX A4000](https://gpucompare.com/gpu/nvidia-rtx-a4000) | $0.13/hr | $3.52/hr | 4 | 7 |

## GPU Models by Vendor







### Other (17 models)
- [NVIDIA B200](https://gpucompare.com/gpu/nvidia-b200)
- [NVIDIA H200 SXM5](https://gpucompare.com/gpu/nvidia-h200-sxm5)
- [NVIDIA A100 80GB SXM4](https://gpucompare.com/gpu/nvidia-a100-80gb-sxm4)
- [NVIDIA H100 SXM5](https://gpucompare.com/gpu/nvidia-h100-sxm5)
- [NVIDIA RTX 6000 Ada](https://gpucompare.com/gpu/nvidia-rtx-6000-ada)
- [NVIDIA L40](https://gpucompare.com/gpu/nvidia-l40)
- [NVIDIA RTX A6000](https://gpucompare.com/gpu/nvidia-rtx-a6000)
- [NVIDIA A40](https://gpucompare.com/gpu/nvidia-a40)
- [NVIDIA L40S](https://gpucompare.com/gpu/nvidia-l40s)
- [NVIDIA RTX 5090](https://gpucompare.com/gpu/nvidia-rtx-5090)
- [NVIDIA V100 32GB SXM2](https://gpucompare.com/gpu/nvidia-v100-32gb-sxm2)
- [NVIDIA L4](https://gpucompare.com/gpu/nvidia-l4)
- [NVIDIA A10](https://gpucompare.com/gpu/nvidia-a10)
- [NVIDIA RTX 4090](https://gpucompare.com/gpu/nvidia-rtx-4090)
- [NVIDIA RTX A5000](https://gpucompare.com/gpu/nvidia-rtx-a5000)
- [NVIDIA A16](https://gpucompare.com/gpu/nvidia-a16)
- [NVIDIA RTX A4000](https://gpucompare.com/gpu/nvidia-rtx-a4000)

## Popular GPU Details

### NVIDIA B200

- **URL**: [https://gpucompare.com/gpu/nvidia-b200](https://gpucompare.com/gpu/nvidia-b200)
- **Slug**: `nvidia-b200`
- **Price range**: $6.52/hr – $48.00/hr
- **Available from**: Hyperstack, Shadeform, Vast.ai, Sesterce Cloud
- **Total offers**: 6

### NVIDIA H200 SXM5

- **URL**: [https://gpucompare.com/gpu/nvidia-h200-sxm5](https://gpucompare.com/gpu/nvidia-h200-sxm5)
- **Slug**: `nvidia-h200-sxm5`
- **Price range**: $3.44/hr – $43.16/hr
- **Available from**: Shadeform, Massed Compute, Vast.ai, Sesterce Cloud, Hyperstack, Nebius
- **Total offers**: 11

### NVIDIA A100 80GB SXM4

- **URL**: [https://gpucompare.com/gpu/nvidia-a100-80gb-sxm4](https://gpucompare.com/gpu/nvidia-a100-80gb-sxm4)
- **Slug**: `nvidia-a100-80gb-sxm4`
- **Price range**: $0.73/hr – $26.40/hr
- **Available from**: Vast.ai, Hyperstack, Shadeform, Massed Compute, Sesterce Cloud, Cudo Compute, Lambda, Vultr
- **Total offers**: 57

### NVIDIA H100 SXM5

- **URL**: [https://gpucompare.com/gpu/nvidia-h100-sxm5](https://gpucompare.com/gpu/nvidia-h100-sxm5)
- **Slug**: `nvidia-h100-sxm5`
- **Price range**: $1.92/hr – $30.80/hr
- **Available from**: Cudo Compute, Vast.ai, Hyperstack, Shadeform, Massed Compute, Sesterce Cloud, Scaleway, Latitude, Nebius
- **Total offers**: 40

### NVIDIA L40

- **URL**: [https://gpucompare.com/gpu/nvidia-l40](https://gpucompare.com/gpu/nvidia-l40)
- **Slug**: `nvidia-l40`
- **Price range**: $0.88/hr – $8.00/hr
- **Available from**: Shadeform, Sesterce Cloud, Hyperstack
- **Total offers**: 13

## Provider Statistics

| Provider | Offers | GPU Models | Regions | Starting From |
|----------|--------|------------|---------|---------------|
| [Sesterce Cloud](https://gpucompare.com/providers/sesterce-cloud) | 89 | 16 | 42 | $0.17/hr |
| [Shadeform](https://gpucompare.com/providers/shadeform) | 89 | 16 | 42 | $0.15/hr |
| [Vast.ai](https://gpucompare.com/providers/vastai) | 47 | 12 | 26 | $0.13/hr |
| [Massed Compute](https://gpucompare.com/providers/massed-compute) | 18 | 6 | 4 | $0.57/hr |
| [Hyperstack](https://gpucompare.com/providers/hyperstack) | 14 | 8 | 4 | $1.50/hr |
| [Verda](https://gpucompare.com/providers/verda) | 13 | 5 | 3 | $0.00/hr |
| [Vultr](https://gpucompare.com/providers/vultr) | 12 | 2 | 9 | $0.47/hr |
| [Nebius](https://gpucompare.com/providers/nebius) | 11 | 3 | 2 | $1.55/hr |
| [Scaleway](https://gpucompare.com/providers/scaleway) | 10 | 3 | 3 | $5.73/hr |
| [Lambda](https://gpucompare.com/providers/lambda) | 6 | 2 | 4 | $1.29/hr |
| [Salad](https://gpucompare.com/providers/salad) | 4 | 3 | 1 | $0.09/hr |
| [Cudo Compute](https://gpucompare.com/providers/cudo-compute) | 3 | 3 | 1 | $6.00/hr |
| [Latitude](https://gpucompare.com/providers/latitude) | 3 | 1 | 3 | $47.98/hr |
| [Novita](https://gpucompare.com/providers/novita) | 3 | 2 | 2 | $0.67/hr |

## Data Available Per GPU Page

Each GPU detail page (e.g., https://gpucompare.com/gpu/nvidia-h100-sxm5) includes:
- Full hardware specifications: VRAM, memory bandwidth (TB/s), TDP (watts), FP32/FP16/FP8 TFLOPS, architecture, form factor
- Pricing overview chart across providers
- Best offers sorted by price
- Provider comparison for that specific GPU
- Schema.org Product JSON-LD with AggregateOffer
- **Markdown version**: `https://gpucompare.com/gpu/nvidia-h100-sxm5.md` — pure text with specs and pricing table

## Data Available Per Provider Page

Each provider page (e.g., https://gpucompare.com/providers/sesterce-cloud) includes:
- Price trend charts over time
- GPU model distribution breakdown
- Complete listing of all offers with pricing
- Regional availability
- Schema.org Organization JSON-LD
- **Markdown version**: `https://gpucompare.com/providers/sesterce-cloud.md` — pure text with GPU models and offers table

## Filters & Search Parameters

The homepage supports filtering by:
- **GPU model**: Select one or more GPU models
- **Provider**: Filter by specific cloud provider(s)
- **Price range**: Set minimum and maximum hourly price
- **GPU count**: Filter by number of GPUs per instance (1x, 2x, 4x, 8x, etc.)

## Machine-Readable Resources

- [Sitemap](https://gpucompare.com/sitemap.xml): XML sitemap with all pages
- [Robots](https://gpucompare.com/robots.txt): Crawler directives (LLM bots explicitly allowed)
- [LLMs.txt](https://gpucompare.com/llms.txt): Summary version of this file
- [LLMs-full.txt](https://gpucompare.com/llms-full.txt): This file (detailed version)