Offers

Browse available GPU instance types with pricing, specifications, and region availability.

The GPU Offers API lets you discover available GPU cluster types on the Huddle GPU marketplace. Offers are real-time snapshots of what's available to deploy right now — including hardware specs, pricing, and which regions have capacity.

Base URL: https://gpu.huddleapis.com/api/v1

What is an Offer?

An offer represents a GPU instance type you can deploy. Each offer includes:

Hardware specs — GPU model, VRAM, vCPUs, RAM, storage
Pricing — per-hour and per-second rates in USD
Availability — which regions currently have capacity
Cluster type — the identifier you use when deploying (e.g., 1H100.80S.30V)

Understanding Cluster Type Names

Cluster types follow a naming convention:

Example	Meaning
`1H100.80S.30V`	1x H100 GPU, 80GB SSD, 30 vCPUs
`8H200.141S.176V`	8x H200 GPUs, 141GB SSD, 176 vCPUs
`1A100.22V`	1x A100 GPU, 22 vCPUs

The number prefix is the GPU count, followed by the GPU model and resource configuration.

List Available GPUs

curl -H "X-API-Key: mk_your_key" \
  https://gpu.huddleapis.com/api/v1/gpus/available

Response:

{
  "code": "OK",
  "data": [
    {
      "gpu_model": "H100",
      "instance_type": "1H100.80S.30V",
      "gpu_count": 1,
      "gpu_vram_gb": 80,
      "total_vram_gb": 80,
      "vcpu_count": 30,
      "ram_gb": 120,
      "storage_gb": 200,
      "price_per_hr": 1.71,
      "price_per_sec": 0.000475,
      "currency": "USD",
      "regions": ["FIN-01", "FIN-03"],
      "is_available": true
    }
  ],
  "meta": {
    "total": 15,
    "has_more": false
  }
}

Filtering and Sorting

Use query parameters to narrow results:

Parameter	Type	Description	Example
`gpu_model`	string	Filter by GPU model	`gpu_model=H100`
`min_vram`	number	Minimum total VRAM in GB	`min_vram=80`
`max_price`	number	Maximum price per hour in USD	`max_price=5.00`
`sort_by`	string	Sort field (`price`, `vram`, `gpu_count`)	`sort_by=price`
`sort_order`	string	Sort direction (`asc`, `desc`)	`sort_order=asc`

# Find all H100 offers sorted by price
curl -H "X-API-Key: mk_your_key" \
  "https://gpu.huddleapis.com/api/v1/gpus/available?gpu_model=H100&sort_by=price&sort_order=asc"

# Find GPUs with at least 80GB VRAM under $5/hr
curl -H "X-API-Key: mk_your_key" \
  "https://gpu.huddleapis.com/api/v1/gpus/available?min_vram=80&max_price=5.00"

Availability Summary

For a quick overview of what's available without full offer details, use the summary endpoint:

curl -H "X-API-Key: mk_your_key" \
  https://gpu.huddleapis.com/api/v1/gpus/available/summary

This returns aggregate stats — available GPU models, price ranges, VRAM ranges, and total offer counts. Useful for building dashboards or quick availability checks.

Prices update in real-time

Offers are synced from the GPU marketplace regularly. Always check current pricing before deploying — prices and availability can change at any time.

Next Steps

Found a GPU you want? Head to Deployments to launch it
GPU not available? Join the Waitlist to get notified when capacity opens up
Check GPU Billing to understand how pricing translates to your credit balance