Easy access to GPU pricing data for major cloud providers: AWS, Azure, GCP, etc. The catalog includes details about prices, locations, CPUs, RAM, GPUs, and spots (interruptible instances).
Usage
import gpuhunt
items = gpuhunt.query(
min_memory=16,
min_cpu=8,
min_gpu_count=1,
max_price=1.0,
)
print(*items, sep="\n")
List of all available filters:
provider: name of the provider to filter by. If not specified, all providers will be used. One or manycpu_arch: CPU architecture, one of:x86,armmin_cpu: minimum number of CPUsmax_cpu: maximum number of CPUsmin_memory: minimum amount of RAM in GBmax_memory: maximum amount of RAM in GBmin_gpu_count: minimum number of GPUsmax_gpu_count: maximum number of GPUsgpu_vendor: GPU/accelerator vendor, one of:nvidia,amd,google,intelgpu_name: name of the GPU to filter by. If not specified, all GPUs will be used. One or manymin_gpu_memory: minimum amount of GPU VRAM in GB for each GPUmax_gpu_memory: maximum amount of GPU VRAM in GB for each GPUmin_total_gpu_memory: minimum amount of GPU VRAM in GB for all GPUs combinedmax_total_gpu_memory: maximum amount of GPU VRAM in GB for all GPUs combinedmin_disk_size: minimum disk size in GB (not fully supported)max_disk_size: maximum disk size in GB (not fully supported)min_price: minimum price per hour in USDmax_price: maximum price per hour in USDmin_compute_capability: minimum compute capability of the GPUmax_compute_capability: maximum compute capability of the GPUspot: ifFalse, only ondemand offers will be returned. IfTrue, only spot offers will be returned
Advanced usage
Every provider catalog is published and versioned independently, so a provider that fails
to be collected keeps its previous catalog while the other providers are updated. Passing
a version to load() requests that version from every provider:
from gpuhunt import Catalog
catalog = Catalog()
catalog.load(version="20260818-42")
items = catalog.query()
print(*items, sep="\n")
Supported providers
- AWS
- Azure
- CloudRift
- Verda
- GCP
- JarvisLabs
- LambdaLabs
- Nebius
- OCI
- Runpod
- Vast AI
- Vultr
See also
2 maintainers
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file gpuhunt-0.1.30.tar.gz.
File metadata
- Download URL: gpuhunt-0.1.30.tar.gz
- Upload date:
- Size: 80.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via:
uv/0.12.5 {"installer":{"name":"uv","version":"0.12.5","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
7ec5a412fca1a4e1b78611fe857eab7efc24027fad4c6d388001c61459fdd655
|
|
| MD5 |
ff4883c8917d6c23308dcbf6d57362a1
|
|
| BLAKE2b-256 |
d6d29e2af6f2482588baa596016171adcdaa6b3c3a72d8ad33bbf41f70d00c09
|
File details
Details for the file gpuhunt-0.1.30-py3-none-any.whl.
File metadata
- Download URL: gpuhunt-0.1.30-py3-none-any.whl
- Upload date:
- Size: 102.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via:
uv/0.12.5 {"installer":{"name":"uv","version":"0.12.5","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
bb8bbdfce0190d80acdbef51658881c8f6bf17b819de40fde8436a75c42d0326
|
|
| MD5 |
69f90fe4d1b47bb2e5f7aa736dade69f
|
|
| BLAKE2b-256 |
034b844578b0ed17176c06cdb3a57f1d2c129139e280f9b94bedf072c862bd38
|
