NVIDIA Model Optimizer is a library that changes deep-learning models through techniques such as quantization, pruning, distillation, neural architecture search, speculative decoding, and sparsity so they can run more efficiently. Developers use it to optimize Hugging Face, PyTorch, or ONNX models and export checkpoints for inference frameworks such as SGLang, TensorRT-LLM, TensorRT, or vLLM. The catalogue entries provide skills, instructions, and a plugin for using its workflows.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add NVIDIA/Model-Optimizer --skill benchmark-model-kernelsgit clone --depth 1 https://github.com/NVIDIA/Model-OptimizerWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/nvidia/model-optimizer/benchmark-model-kernels)<a href="https://agentmods.dev/skills/nvidia/model-optimizer/benchmark-model-kernels"><img src="https://agentmods.dev/badge/skills/nvidia/model-optimizer/benchmark-model-kernels/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nvidia/model-optimizer/benchmark-model-kernels"><img src="https://agentmods.dev/badge/skills/nvidia/model-optimizer/benchmark-model-kernels.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00105 | $0.01962 |
| Opus 5.5 | $0.00042 | $0.00785 |
| Sonnet 5.5 | $0.00021 | $0.00392 |
| Haiku 4.5 | $0.00011 | $0.00196 |
Grade A, and why
benchmark-model-kernels scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Benchmark Model Kernels
Plan a single-GPU microbenchmark for the per-rank shapes of an intended deployment. The scripts never load checkpoint weights, launch distributed workers, or measure collectives, serving throughput, or request latency.
Interview, one decision per message
Ask exactly one unresolved decision per message with two or three concrete options, state the default, and wait for the answer. Recommend an option only when it is substantively better. Skip decisions already answered. Follow this order:
-
Model — a local directory, a
config.json, or a Hub ID (equivalent; no default). The script builds the model on meta tensors from configuration only. Do not enable--trust-remote-codewithout explicit approval; pin--revision <commit>when it is needed. -
TP and EP — accept both directly, or derive them from the intended deployment's GPU model and count and confirm. Defaults:
TP=1,EP=1. Derive parallelism from the intended deployment, never from the GPU that runs the microbenchmark. The script validates every sharding rule (divisibility, GQA replication, expert partitioning) and errors loudly. -
M sweep — balanced default (recommended):
1 8 64 512; decode-focused:1 4 16 32; throughput-focused:64 256 1024 4096. M is roughly the tokens scheduled per step (decode: active sequences), not endpoint concurrency. With EP, an MoE row models one rank's share of the global batch: a global batch of B tokens corresponds to the column M = B/EP, so do not compare different EP values at the same M. -
Shape preview — always run this before anything else; no FlashInfer or GPU needed:
python "$SKILL_DIR/scripts/benchmark_model.py" <model> \ --tp <tp> --ep <ep> --ms <m1> <m2> ... --print_onlyReview the printed shapes and the MoE tuple with the user.
# unsupported:lines mean the list is partial and the script exits nonzero — handle those via Manual supplements below. -
FlashInfer checkout and GPU — the full benchmark needs a FlashInfer source checkout containing
benchmarks/flashinfer_benchmark.py; the installed wheel alone is not enough. Prefer a clean checkout matching the installedflashinferversion; ask before cloning or installing anything. On the benchmark machine, checknvidia-smiand package versions, verify the target GPU is idle (concurrent work on the same GPU skews timings), and verify CUPTI timing with a tinybench_gpu_time(..., enable_cupti=True)probe — a warning that falls back to CUDA events is a failure (thecupti-python/nvidia-cuda-cuptipackages must match PyTorch's CUDA major). Ask for a GPU index only when several are visible. Pick a fresh workdir and state it. -
Full benchmark:
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 146 lines · 105 tokens per session scan A d1e3238ea29c
benchmark-model-kernels is a skill published in the GitHub repository NVIDIA/Model-Optimizer (5,398 stars, last pushed today), licensed Apache-2.0. It adds 105 tokens to every session and 1,962 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
torchdrug
Builds and troubleshoots TorchDrug 0.2.1 workflows for molecular graphs, property prediction, self-supervised pretraining, molecule generation, retrosynthesis, protein representation learning, and knowledge graph reasoning. Use when code imports torchdrug or needs its datasets, models, tasks, or Engine.
arboreto
Infers candidate gene regulatory networks from bulk or single-cell expression data using AertsLab Arboreto GRNBoost2 and GENIE3. Use for transcription factor-target association ranking, compatible Dask execution, sparse expression inputs, and network stability checks.
deepspot-m
Generates transcriptome-wide virtual spatial transcriptomics from H&E histology with DeepSpot-M. Used for predicted log1p-CPM expression from 224x224 tiles at about 20x, querying the released protein-coding gene panel by symbol, and whole-slide prediction after resolution-aware tiling with histolab.
pyhealth
Builds and validates PyHealth clinical machine-learning pipelines for EHR, signals, imaging, and medical codes. Use for PyHealth dataset loading, MIMIC-III/IV, eICU or OMOP prediction tasks, patient-level evaluation, mortality/readmission/length-of-stay modeling, medication recommendation, sleep staging, Trainer…
ml-training-recipes
Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics, genomics. Covers training loops, optimizer selection (AdamW, Muon), LR scheduling, mixed precision, debugging, and systematic experimentation. Use when training or fine-tuning…
nemo-mbridge-perf-expert-parallel-overlap
Validate and use MoE expert-parallel communication overlap in Megatron-Bridge, including overlapmoeexpertparallelcomm, delaywgradcompute, and flex dispatcher backends such as DeepEP and HybridEP.