providers
offerings
4
GPU SKUs
2
regions
us · eu
tags
inference · training · open-models
:: GPU compute SKUs
GPUVRAMregion$/hr$/gpu·secnotesverified
- A100-80GB[cheapest]80 GB美洲$1.50——2d ago
- H10080 GB美洲$2.40—Dedicated endpoint2d ago
:: offerings · pricing · limits
filter:4 / 4
modelctxstatuspricing · $/Mrate limitsverified
- 128Kactive[标准] in $3.50/M out $3.50/M—2d ago
Llama 3.1 405B
文本Function CallingJSON Mode流式
- 128Kactive[标准] in $0.880/M out $0.880/M—2d ago
Llama 3.3 70B
文本Function CallingJSON Mode流式
- 131Kactive[标准] in $0.200/M out $0.600/M—2d ago
Qwen-Plus
文本图像Function CallingJSON Mode视觉
- 33Kactive[标准] in $0.300/M out $0.900/M—2d ago
QwQ-32B (Preview)
文本推理