| GPU model | VRAM | Current Lowest Price | Platform | Per day | Generate a batch of short videos (6-hour estimate) | |
|---|---|---|---|---|---|---|
| RTX 4090 All-rounder workhorse: Flux image generation, QLoRA fine-tuning of 13B models, video generation. | 24GB | ¥0.87/hour | Vast.ai (USA) Spot | ¥21 | ¥5 | Price comparison → |
| RTX 5090 The RTX 4090 successor with larger VRAM enables smoother video generation and fine-tuning of 13B models | 32GB | ¥1.53/hour | Vast.ai (USA) Spot | ¥37 | ¥9 | Price comparison → |
| A100 80GB Full-parameter fine-tuning for 7B models; quantized inference for 70B models; medium-scale training tasks | 80GB | ¥5.47/hour | Paperspace (United States) Pay-as-you-go | ¥131 | ¥33 | Price comparison → |
| L40S 48GB 48 GB VRAM sweet-spot GPU: ideal for 70B quantized inference and batch inference | 48GB | ¥5.69/hour | RunPod (USA) Spot | ¥137 | ¥34 | Price comparison → |
Start with RTX 4090 (24GB); for long or high-resolution video move to RTX 5090 / L40S 48GB / A100 80GB. Insufficient VRAM is the most common cause of failure.
Hourly rate × actual generation time. This page shows a 6-hour reference total; the real figure depends on resolution and frame count.
Video generationCost估算｜Data Source：GPU computing power radar rentgpu.link（2026-10-26 更新，per hour自动抓取各Platform公开定价）｜https://rentgpu.link/use/video-gen｜CC BY 4.0
rentgpu.link
Neutral GPU compute rental price comparison tool: collects publicly listed prices from 26 global platforms every hour. We do not own GPUs nor sell compute.