Choose a cloud GPU by workload

Start with the application’s memory and software requirements, then compare rental prices. These guides do not claim measured performance.

• Best Cloud GPU to Rent for 3D Rendering$0.35/hr, RTX 4090GPU rendering rewards raw ray-tracing throughput, so the value pick is an RTX 4090 at $0.35/hr, which renders faster per dollar than datacenter cards in engines like Blender Cycles, Octane, and Redshift. Compares RTX 5090, RTX 4090, RTX A6000 and L40S.
• Cloud GPU for self-hosted AI agents$0.44/hr, L4Price the model-serving workload separately from the agent’s tools and orchestration. Compares L4, L40S, A100 and H100.
• Cloud GPU for batch inference$0.35/hr, RTX 4090Choose the lowest total compute cost among configurations that meet your completion deadline and model memory requirement. Compares RTX 4090, L40S, A100 and H100.
• Cloud GPU for CAD and remote workstations$0.35/hr, RTX A6000Check your CAD application’s supported GPU, driver, operating system and remote-display licensing before choosing a cloud workstation. Compares RTX A6000, RTX 6000 Ada and RTX PRO 6000.
• Cloud GPU for cloud gaming and game streaming$0.35/hr, RTX 4090A cloud GPU compute offer is not automatically a gaming service. Compares RTX 4090, RTX 5090 and RTX 6000 Ada.
• Best Cloud GPU for Computer Vision$0.35/hr, RTX 4090Most computer vision work, from detection and segmentation to classification, fits comfortably on a 24GB RTX 4090 at $0.35/hr, which is the value pick for both training and batch inference. Compares RTX 4090, L4, L40S and A100 80GB.
• Cloud GPU for text embeddings$0.35/hr, RTX 4090Select hardware around the embedding model, input length and batch size. Compares L4, RTX 4090 and A100.
• Best Cloud GPU for Jupyter Notebooks$0.35/hr, RTX 4090For interactive Jupyter work the goal is a cheap, always-ready card you can spin up and tear down, so the value pick is an RTX 4090 at $0.35/hr: 24GB covers most experimentation, prototyping, and small training. Compares RTX 4090, L4, RTX 5090 and A100 80GB.
• Best Cloud GPU for LLM Fine-Tuning$0.35/hr, RTX 4090For LoRA and QLoRA fine-tunes of 7B to 13B models the value pick is an RTX 4090 at $0.35/hr: 24GB is enough once weights are quantized and only adapters are trained. Compares RTX 4090, A100 80GB, H100 and H200.
• Best Cloud GPU for LLM Inference$0.35/hr, RTX 4090Rent the cheapest card whose memory holds your model and its context: a 24 GB RTX 4090 at $0.35/hr serves 7B to 13B models at 4-bit, a 48 GB L40S at $0.96/hr holds a 4-bit 70B for one user at moderate context, and an 80 GB A100 ($1.09/hr) or H100 ($1.99/hr) leaves room for long context and many users. Compares RTX 4090, L40S, A100 80GB and H100.
• Best Cloud GPU for LLM Training$1.09/hr, A100Training or continued pretraining of an LLM is a multi-GPU job where interconnect and memory bandwidth decide cost, so the value pick is an 80GB A100 at $1.09/hr and the performance pick is an H100 at $1.99/hr. Compares A100 80GB, H100, H200 and B200.
• Cloud GPU for GPU simulation$0.89/hr, A100Start with the solver’s supported hardware and numerical precision. Compares A100, H100 and MI300X.
• Cloud GPU for speech recognition$0.35/hr, RTX 4090Budget for the model, audio duration and concurrency you will actually process. Compares L4, RTX 4090 and A100.
• Best Cloud GPU to Rent for Stable Diffusion & ComfyUI$0.35/hr, RTX 4090For most Stable Diffusion work the value pick is an RTX 4090 at $0.35/hr: 24GB of VRAM runs SDXL comfortably and its raw image throughput beats far pricier datacenter cards. Compares RTX 4090, RTX 5090, L40S and A100 80GB.
• Best Cloud GPU for AI Video Generation$0.35/hr, RTX 4090AI video models are memory-hungry because they generate many frames at once, so the sensible floor is 48GB and the value pick is an L40S, with an 80GB A100 at $1.09/hr or H100 at $1.99/hr for longer clips and higher resolution. Compares L40S, A100 80GB, H100 and RTX 4090.

For a named language model, use the model memory pages. For a fixed budget, start with the GPU price table.