The most thorough AI hardware directory online. Browse 109 devices, data center GPUs (H100, MI300X, B200), Apple Silicon, AMD Radeons, Google TPUs, edge devices like Jetson and Coral, AI PCs, and humanoid robots including Optimus, Figure, and Unitree. Every listing shows VRAM, TFLOPS, price, and availability at a glance.

Filter by manufacturer, use-case tags, VRAM range, or price. Compare up to 3 products side by side. Each product has its own dedicated page with metric explainers, AI video previews, images, and links. Download the table, or embed it anywhere. Subscribe to drop alerts when new hardware is added.

52 open-source models ranked by a composite score built from real benchmarks: reasoning, code, math, vision, multilingual, function calling, and more. Every model shows its parameter count, context window, Q4 VRAM footprint, and which hardware tiers it runs on (8 GB / 16 GB / 32 GB / 64 GB+).

Filter by capability, architecture (Dense vs MoE), parameter count, or your VRAM budget. Pick a closed model like GPT-5.4 and it recommends the best open-source alternatives. Each model has its own detail page with benchmark breakdowns and compatible hardware. Benchmark data is updated continuously via Hugging Face API.

Stop guessing. The two tools below answer the two questions every AI practitioner eventually asks: "Can my hardware actually run this model?" and "Is it cheaper to buy hardware or keep paying for API access?" Both run entirely in your browser, are fully embeddable, and have a share button so you can send results to your team.

Enter your VRAM (or auto-detect your system), choose model sizes, quantization level (Q2 through FP16), and architecture, and see exactly which AI models your hardware can run, how fast, and at what quality. It accounts for quantization trade-offs, MoE vs dense architectures, and context window requirements.

The calculator is also fully embeddable, copy a single snippet and drop it into your blog or documentation so your readers can check their own specs without leaving your page.

The buy-vs-API question answered in numbers. Select your hardware, pick an API model to compare against, then dial in your usage: tokens per day, hours of use, input/output ratio, electricity cost, and time horizon. Quick presets cover Casual User, Developer, Power User, and Enterprise.

The calculator shows your break-even point and total savings over 36 months, factoring in hardware cost, power draw, and API pricing. Also embeddable and shareable, paste your results link directly into a Slack thread or slide deck.