Intel GPU AI Skills is a collection of agent skills for setting up, running, benchmarking, and profiling Hugging Face models on Intel GPUs. It supports workflows involving PyTorch, vLLM-XPU, SGLang-XPU, llama.cpp-SYCL, and migration from CUDA to XPU. The catalogue contains the project's skills, instructions, agent, and plugin.
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 intel/gpu-ai-skills --skill model-can-it-fitgit clone --depth 1 https://github.com/intel/gpu-ai-skillsWrote 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/intel/gpu-ai-skills/model-can-it-fit)<a href="https://agentmods.dev/skills/intel/gpu-ai-skills/model-can-it-fit"><img src="https://agentmods.dev/badge/skills/intel/gpu-ai-skills/model-can-it-fit/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/intel/gpu-ai-skills/model-can-it-fit"><img src="https://agentmods.dev/badge/skills/intel/gpu-ai-skills/model-can-it-fit.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 | $0.00106 | $0.02341 |
| Opus 5 | $0.00053 | $0.01171 |
| Sonnet 5 | $0.00021 | $0.00468 |
| Haiku 4.5 | $0.00011 | $0.00234 |
Grade A, and why
model-can-it-fit 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 10d 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 — 213 lines — stays where its author put it; the contents beside it link to each section on GitHub.
model-can-it-fit
Use this for a pre-launch VRAM calculator: whether a Hugging Face model can fit on an Intel GPU at a requested quantization, context length, concurrency, runtime, and tensor-parallel degree. Input: HF model id, quantization, context length, concurrency, target VRAM. Output: a per-component breakdown and a verdict.
The skill runs a CPU-only calculator. It does not need to deploy the model on an Intel GPU, but it MUST check the Intel GPU VRAM memory space and may need Hub access unless the user provides a local config.json or params.json.
Use And Route
Use this skill when the user asks:
- whether a model fits on an Intel GPU or XPU
- what max context or concurrency is memory-feasible
- how VRAM splits across weights, KV cache, activations, and runtime
- whether a vLLM/SGLang launch is likely to OOM before trying it
Use another skill instead when the user asks for:
- measured speed, TTFT, TPOT, or tokens/sec: use a benchmark skill
- an exact launch configuration or performance recommendation: use
model-config-recommend - diffusion fit: use
torch-xpu-benchempirically - live GPU readiness: use
xpu-runtime-preflightorxpu-discover
Inputs To Collect
Ask for or infer:
- model id or local config path
- target GPU VRAM per device -- if the target is this host, measure it (see "Measure VRAM first" below) instead of asking the user
- runtime:
vllm,sglang, ortorch - quantization:
bf16,fp16,fp8,int8,int4,int3,int2, ormxfp4 - context length and concurrency
- tensor parallel degree if multiple XPUs are planned
- vLLM
--gpu-memory-utilizationvalue if this is launch planning
For gated Hugging Face repos, use HF_TOKEN or
HUGGING_FACE_HUB_TOKEN. For repeatable tests, prefer local config
snapshots.
Measure VRAM First
--device-vram-gb is required and has no default. Confirm which card
the host actually has before choosing a value -- never assert VRAM from a
remembered spec sheet:
xpu-smi discovery -d 0 | grep -i 'Device Name\|Memory Physical Size'
What ships with it
3 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.
- 10d ago First seen · 213 lines · 106 tokens per session scan A 1a7f87af72a9
model-can-it-fit is a skill published in the GitHub repository intel/gpu-ai-skills (21 stars, last pushed 5d ago), licensed Apache-2.0. It adds 106 tokens to every session and 2,341 once invoked, about $0.0005 per session on Opus 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
spark-environment-setup
Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.
spark-memory-thermal-ops
Manage unified memory and thermals during long-running ML jobs on NVIDIA DGX Spark. Use when planning memory headroom for a training run on GB10, when a job OOMs on unified memory, or when monitoring temperature and power during multi-hour training.
spark-training-gotchas
Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.
llama-cpp
Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M3 Macs, AMD/Intel GPUs, or when CUDA is unavailable. Supports GGUF quantization (1.5-8 bit) for reduced memory and 4-10× speedup vs PyTorch on CPU.
minicpm5-deploy-vllm-ascend
Deploy MiniCPM5-2B with vLLM on Huawei Ascend NPU using vLLM-Ascend. Use when the user mentions vLLM-Ascend, Ascend NPU, Huawei Ascend, CANN, torchnpu, davinci devices, or wants an OpenAI-compatible MiniCPM5 server on Ascend hardware.
amc-run-rtsp-calibration
Calibrate a new dataset from live RTSP camera streams via the AutoMagicCalib REST API. Use when the user provides RTSP URLs or asks to calibrate live cameras; VIOS records clips, AMC ingests them, then runs calibration.