aws-samples/sample-gpu-open-weight-bench (bench-open-model)

Benchmark any open-weight vLLM-servable Hugging Face model on AWS GPU hardware. Sizes the model's VRAM needs, picks the right EC2 GPU instance type and region, provisions it, serves the model on vLLM, runs a cache-honest concurrency sweep, writes a report, and tears everything down. Use when asked to benchmark / test / evaluate the throughput, latency, or serving capacity of an open-weight model on AWS, pick a GPU instance type for a model, size VRAM for a model, or measure tokens-per-second or requests-per-day. Covers text LLMs and multimodal / vision / OCR models.

Claude Skills are markdown + scripts loaded into Claude Code; scanned with the same pipeline.

Source:

github.com/aws-samples/sample-gpu-open-weight-bench

Provenance:

Found by crawling (provenance, not a security rating)

We found this by crawling public sources and nobody has attested to it. That is how most of the registry gets here, and it is not a mark against the artifact — it only means no authenticated act of publication is on record. An artifact can be the vendor's own and still appear here.

3 critical findings reported by cisco-skill-scanner. The composite score alone does not reflect this — read the per-scanner reports below.