Thunder Compute
On-demand cloud GPUs for AI and ML work, with a CLI and a VS Code and Cursor extension.
Introduction
Thunder Compute rents on-demand cloud GPU instances, including T4, A100, and H100 models, billed by the hour. Users can start instances from templates such as Ollama, use a command-line tool, and sync files in real time from VS Code or Cursor. The service says there are no long-term commitments, and new users can start with free credit.
- Added on
- Nov 2, 2026
- Website
- thundercompute.com
- Pricing model
- Paid
Product information
What is Thunder Compute?
Thunder Compute is a cloud GPU service for training, fine-tuning, and running AI models.
How to use Thunder Compute?
Create an instance, choose a model or template, and run it. Use the VS Code or Cursor extension to edit files on the instance.
Core features of Thunder Compute
- On-demand A100 and H100 instances
- Simple CLI
- Templates for common AI tools
- VS Code and Cursor sync extension
- Pay-as-you-go billing
Use cases of Thunder Compute
- 1Prototyping AI and ML projects
- 2Training and fine-tuning models
- 3Self-hosting AI applications
- 4Transcribing video
Frequently asked questions about Thunder Compute
Which GPUs can I use on Thunder Compute?+
Instances include Tesla T4, A100 (40GB and 80GB), and H100 GPUs.
How is Thunder Compute billed?+
Usage is charged by the hour with no commitment. The site says new users can start with $20 in free credit.
Where are Thunder Compute's servers located?+
The listing says the servers are in the U.S. Central region.
Which category is Thunder Compute in?+
Thunder Compute is listed under AI Code Editor.