Parameters
MoE · FP8 · 160 GBOpen Model Catalogue
DeepSeek
Text generationReasoningCoding
DeepSeek V4 Flash
deepseek-ai/DeepSeek-V4-FlashDeepSeek newest, compact — 2 × H200
Min VRAM
≈ 190 GBHardware
3 × 80 GB GPUOpenAI-compatible API with a per-workload key — drop-in for existing SDKs.
Dedicated GPUs — your model, your instance, no shared queues.
Billed by the second with auto-shutdown — never pay for idle time you didn't plan.