How do I authenticate the RunPod stock skill?

The skill requires a RunPod API key, which can be provided via the RUNPOD_API_KEY environment variable or a ~/.runpod/config.toml file.

Is the stock status 100% accurate?

The skill retrieves data from the RunPod API. However, because the API may sometimes return 'High' stock regardless of actual availability, it is recommended to verify via the RunPod Web UI before starting a deployment.

Can I find GPUs that support persistent network volumes?

Yes, the skill includes a --storage filter that restricts results to datacenters where Network Volumes (S3) are supported.

Does this skill provide real-time pricing for RunPod instances?

Yes, it displays current pricing for GPU instances, defaulting to Community Cloud rates and falling back to Secure Cloud if needed.

Can I filter for specific GPU architectures like NVIDIA Hopper?

Yes, you can use the --gen flag to filter specifically for architectures such as Blackwell, Hopper, Ada, Ampere, Volta, or AMD.

RunPod GPU Stock Checker

Name: RunPod GPU Stock Checker
Author: pokutuna

bypokutuna

•

Cloud Infrastructure

Monitors and queries real-time RunPod GPU availability, specifications, and pricing to optimize compute resource allocation.

The RunPod GPU Stock Checker skill empowers developers and data scientists to locate available high-performance compute resources directly through Claude. By interfacing with the RunPod API, it allows for granular searches across global datacenters, filtering by VRAM, GPU generation (such as Blackwell, Hopper, or Ada), and specific hardware models. It is particularly useful for identifying cost-effective instances for ML training or inference, providing clear visibility into stock levels, network volume support, and pricing across both Secure and Community clouds.

Key Features

01Query specific VRAM requirements (e.g., 80GB+) and GPU models

02Compare pricing between Community and Secure cloud instances

03Filter GPUs by generation including Blackwell, Hopper, and Ampere

04View stock levels and location data for global GPU availability

051 GitHub stars

06Identify datacenters with Network Volume (S3) support

Use Cases

01Checking availability of the latest RTX 5090 or 4090 GPUs for cost-effective inference

02Finding available H100 or H200 instances for intensive machine learning training sessions

03Locating datacenters with specific storage capabilities to match persistent volume requirements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add pokutuna/claude-plugins stocks

For use in Claude.ai and ChatGPT

Download Skill