InfraSep 23, 2026 · 19:402 sources

vLLM Adds DiffusionGemma-Jev to Ease Cloud AI Deployment

From marketbrief.now

Image: marketbrief.now
AI;DR

The vLLM project announced on September 23 that its engine now supports DiffusionGemma-Jev, enabling the model to return confidence scores by reading from noisy answer slots. Google offers one-command deployment of the model on Cloud Run, providing single-step latency between 35 and 60 ms and throughput of 100 to 123 requests per second for batch sizes of 32. Hosting through Google Cloud Run costs approximately $3 per hour and drops to $0 when the system is idle.

Read the sourceSign in to suggest

Sources

#vllm#diffusiongemma-jev#cloud#cloud-run#api-deployment#deployment#apiMarkdown
Content history (2)
  • 2026-10-16 · Summary · vi · Replaced stored text
    Google hiện cung cấp đường dẫn triển khai chỉ bằng một lệnh cho mô hình DiffusionGemma-Je…
    Google cung cấp đường dẫn triển khai một lệnh cho mô hình DiffusionGemma-Jev trên Cloud R…
  • 2026-10-16 · Title · vi · Replaced stored text
    vLLM Hỗ trợ DiffusionGemma-Jev để Triển khai AI trên Cloud Dễ dàng hơn
    vLLM thêm DiffusionGemma-Jev để dễ triển khai AI trên Cloud

One ranked edition, every morning

The stories that matter today. Free, cancel anytime.

Telegram @aidr_today