About Will It Run AI

The largest free on-device AI compatibility platform — check whether your GPU or Mac can run local AI models with deeper analysis than any alternative.

What we do

Will It Run AI answers one question: "Can my hardware run this AI model?" We go beyond simple VRAM checks. Our engine scores 10 factors including memory bandwidth, decode speed, context capacity, offload ratio, quantization quality, and workload-specific fit to give you an accurate compatibility grade.

We cover 300+ AI models (LLaMA, DeepSeek, Qwen, Mistral, Gemma, Phi, and more) across 195+ hardware profiles including NVIDIA GPUs, AMD GPUs, Intel GPUs, and Apple Silicon Macs.

Our methodology

Every fit estimate is computed from a 10-factor scoring model that considers:

1. 01VRAM/Memory capacity — Does the model fit at all?
2. 02Memory bandwidth — How fast can data flow to the GPU?
3. 03Decode speed — Tokens per second during generation
4. 04Time-to-first-token — Prompt processing latency
5. 05Context capacity — Maximum safe context window
6. 06Offload ratio — How much spills to system RAM?
7. 07Quantization quality — How much fidelity is preserved?
8. 08Workload fit — Optimized for chat, coding, reasoning, or RAG
9. 09Architecture efficiency — MoE, dense, hybrid considerations
10. 10Hardware efficiency — Generation-specific compute factors

Read the full methodology on our Why It Works page.

Data sources

Our catalog is built from authoritative sources and continuously enriched:

• Hugging Face — Model metadata, parameter counts, architecture details, download stats
• Ollama — Distribution channels, runtime compatibility, model availability
• Official documentation — GPU specifications, Apple Silicon specs, MSRP data
• Open LLM Leaderboard — Quality benchmarks and tier assignments

All estimates are approximations. Actual performance depends on software configuration, system load, and runtime implementation. Do not make purchasing decisions based solely on these estimates.

Who we are

Will It Run AI is built by Fase Consulting Ibiza, S.L. (NIF: B57969656), a software consultancy based in Ibiza, Spain. We build tools for developers and AI practitioners who want on-device AI — running models locally rather than relying on cloud APIs.