8341TokenFactory

GPU Compute Marketplace

Hourly billing — you only pay while it runs. Training, inference, video generation and image generation on one machine.

11
GPU types
80 GB
Max VRAM
Hourly
Billing
99.9%
Uptime

GPUs & Pricing

Filter by compute, VRAM and price. Every GPU supports hourly billing — the meter stops when you power off.

11 GPUs available
GPUVRAMFP16 TFLOPSHourlyMonthlyBest forAction
NVIDIA H800
High-End · 大规模训练首选
80 GB756¥9.98 /hr¥5,999 /mo
训练大模型微调
Rent now
NVIDIA RTX PRO 6000
High-End · 视频、图像生成的高性价比选择
96 GB503.8¥8.36 /hr¥4,999 /mo
视频生成图像生成推理
Rent now
NVIDIA A800 80GB
High-End
80 GB312¥5.59 /hr¥3,499 /mo
训练推理
Rent now
NVIDIA RTX 5090
Mainstream
32 GB210¥2.78 /hr¥1,699 /mo
推理个人开发
Rent now
NVIDIA RTX 4090
Mainstream
24 GB165.2¥1.88 /hr¥1,099 /mo
推理微调出图
Rent now
NVIDIA H20
High-End
96 GB148¥9.98 /hr¥5,999 /mo
推理长上下文
Rent now
NVIDIA V100
Mainstream
32 GB125¥1.88 /hr¥1,099 /mo
训练老项目兼容
Rent now
NVIDIA RTX A4000
Entry
16 GB76.7¥0.92 /hr¥559 /mo
轻量推理开发调试
Rent now
NVIDIA RTX 3090
Mainstream
24 GB71¥1.32 /hr¥799 /mo
推理出图
Rent now
NVIDIA RTX 3080Ti
Entry
12 GB70¥0.98 /hr¥599 /mo
轻量推理
Rent now
NVIDIA RTX 2080Ti
Entry
11 GB53.8¥0.88 /hr¥529 /mo
开发调试学生
Rent now

Estimate your cost

Drag the sliders to see what your workload costs. If usage exceeds the monthly plan, we calculate the cheaper option automatically.

Per hour
¥9.98 /hr
Per day
¥79.84
Estimated monthly
¥1,756
On-demand billing

Estimates only — your invoice is the source of truth. Billing stops when powered off.

Will my model fit in VRAM?

Pick model size and precision to see the VRAM needed, plus the cheapest GPU that can run it.

Estimated VRAM needed
18 GB
Recommended GPU
NVIDIA RTX 3090¥1.32 /hr≈ ¥964 /mo

Estimated from weights + optimizer + activation headroom. Actual depends on the implementation.

Compute leaderboard

Ranked by FP16 throughput to help you balance performance against cost.

1
NVIDIA H800
80 GB · ¥9.98/h
756 TFLOPS
2
NVIDIA RTX PRO 6000
96 GB · ¥8.36/h
503.8 TFLOPS
3
NVIDIA A800 80GB
80 GB · ¥5.59/h
312 TFLOPS
4
NVIDIA RTX 5090
32 GB · ¥2.78/h
210 TFLOPS
5
NVIDIA RTX 4090
24 GB · ¥1.88/h
165.2 TFLOPS
6
NVIDIA H20
96 GB · ¥9.98/h
148 TFLOPS
7
NVIDIA V100
32 GB · ¥1.88/h
125 TFLOPS
8
NVIDIA RTX A4000
16 GB · ¥0.92/h
76.7 TFLOPS
9
NVIDIA RTX 3090
24 GB · ¥1.32/h
71 TFLOPS
10
NVIDIA RTX 3080Ti
12 GB · ¥0.98/h
70 TFLOPS
11
NVIDIA RTX 2080Ti
11 GB · ¥0.88/h
53.8 TFLOPS

Values are vendor-published FP16 tensor throughput; real performance also depends on memory bandwidth and model architecture.

How it works

1

Pick a GPU

Choose by VRAM and compute, order hourly or monthly.

2

Boot and go

Deep-learning stack preinstalled, mainstream frameworks ready out of the box.

3

Run your job

Connect via SSH or Jupyter — train, infer, generate.

4

Stop and stop paying

Power off to stop billing. Your data stays, restart anytime.

Do the math before you buy

Not sure about specs? Tell us your model and budget and we'll size it for you.