Amazon EC2
Rent AWS GPU servers, from NVIDIA T4 to B200, and run any open model yourself.
Overview
Amazon EC2 gives you GPU and AI accelerator servers to run any open-weight model yourself, with full control over the software stack. Instances range from a single NVIDIA T4 to eight NVIDIA B200s, billed per instance-hour on demand, with lower prices for Spot capacity, Savings Plans and Capacity Blocks.
Specifications
| Hosting type | GPU cloud |
|---|---|
| Model families | Mistral, DeepSeek, Gemma, gpt-oss, Llama, Qwen |
| LLM models | Codestral 22B, DeepSeek-R1, DeepSeek-V3, DeepSeek-V3.1, Gemma 2 27B, Gemma 2 9B, Gemma 3 12B, Gemma 3 27B, Gemma 3 4B, gpt-oss-120b, gpt-oss-20b, Llama 3.1 8B Instruct, Llama 3.3 70B Instruct, Mistral NeMo, Mistral Small 3.1, Qwen2.5 72B Instruct, Qwen2.5-Coder 32B Instruct, Qwen3 235B-A22B, Qwen3 30B-A3B, Qwen3 32B |
| Pricing range | From US$0.526 |
| Pricing model | Per GPU-hour, Reserved or committed capacity |
| Free tier available | No |
| Data center locations | United States, Canada, Mexico, Brazil, Ireland, United Kingdom, Germany, France, Sweden, Italy, Spain, Switzerland, Israel, Bahrain, United Arab Emirates, South Africa, India, Japan, South Korea, Singapore, Australia, Indonesia, Malaysia, Thailand, Taiwan, Hong Kong SAR China, New Zealand, China |
| Hardware / GPU options | G4dn (NVIDIA T4), G5 (A10G), G6 (L4), G6e (L40S), P4d (A100), P5 (H100), P5e/P5en (H200) and P6 (B200) instances, plus Inf2 (AWS Inferentia2) and Trn1/Trn2 (AWS Trainium). |
| VRAM & memory | 16 GB – 180 GB |
| Maximum model size | 671B |
| OpenAI-compatible API | No |