Ollama
Ollama daemon
Needs 7 GB VRAM
Works on
No GPU online for this yet
Browse GPUs and AI apps
Deploy on a live GPU, or pick an app and we find the GPU. Prices include the 5% fee.
Deploy on a live GPU, or pick an app and we find the GPU. Prices include the 5% fee.
Got a GPU? Run this on it and start earning.
curl -fsSL https://amplerun.com/host | sudo bashOllama daemon
Needs 7 GB VRAM
Works on
No GPU online for this yet
SSH workspace with vLLM installed
Needs 15 GB VRAM
Works on
No GPU online for this yet
faster-whisper speech-to-text behind an OpenAI-compatible API
Needs 3 GB VRAM
Works on
No GPU online for this yet
Chat with an open model
Needs 15 GB VRAM
Works on
No GPU online for this yet
ComfyUI that downloads FLUX.1 [schnell] at start
Needs 15 GB VRAM
Works on
No GPU online for this yet
Fine-tune a model
Needs 22 GB VRAM
Works on
No GPU online for this yet
Chat with an open model
Needs 7 GB VRAM
Works on
No GPU online for this yet
Chat UI with a bundled Ollama
Needs 7 GB VRAM
Works on
No GPU online for this yet
Qwen3 8B, 4-bit GGUF, on llama.cpp
Needs 7 GB VRAM
Works on
No GPU online for this yet
A GPU notebook or dev box
Needs 7 GB VRAM
Works on
No GPU online for this yet
NVIDIA CUDA 12.9 + cuDNN development image with SSH
Needs 3 GB VRAM
Works on
No GPU online for this yet
Chat with an open model
Needs 8 GB VRAM
Works on
No GPU online for this yet
ComfyUI v0.37.4, node-based image and video generation
Needs 7 GB VRAM
Works on
No GPU online for this yet
Qwen3 8B, AWQ 4-bit, on vLLM
Needs 11 GB VRAM
Works on
No GPU online for this yet
ComfyUI that downloads SDXL base 1.0 at start
Needs 7 GB VRAM
Works on
No GPU online for this yet
A GPU notebook or dev box
Needs 7 GB VRAM
Works on
No GPU online for this yet
Fine-tune a model
Needs 8 GB VRAM
Works on
No GPU online for this yet
gpt-oss-20b on vLLM for high-throughput batched serving
Needs 22 GB VRAM
Works on
No GPU online for this yet
Unsloth's fast LoRA/QLoRA fine-tuning in JupyterLab
Needs 15 GB VRAM
Works on
No GPU online for this yet