AI article
EmbeddingGemma 2 on Cloud Run: Minimum Specs, and When an NVIDIA L4 Pays for Itself
Community description: How small a Cloud Run instance EmbeddingGemma 2 fits in, whether an int4 repack shrinks it further, and how many documents a burst needs before an NVIDIA L4 costs less than a 1 vCPU int8 service.
Dev.to | Oct 15, 2026 | xbill
Automated excerpt
Minimum spec: 1 vCPU, 2 GiB, int8 weights. Queries take a median 133 ms, a million documents cost USD 27. 67, and an idle service costs nothing. 1 GiB is too small. The CPU service embeds 27. 6 million documents for the same money.
Selected automatically from source text; not independently written or fact-checked. Read the original for full context.