v1.1.1 — readme catch-up

Documentation patch on top of v1.1.0. README brought in line with the features that landed in v1.1 (Hugging Face URL import, gated config.json paste, share links, markdown export, EN/ES toggle, dark/light theme, comparison chart) and the expanded hardware coverage (NVIDIA GeForce/Datacenter, AMD Radeon, Copilot+ PC NPUs). Adds a user-flow Mermaid diagram and softens hard-coded preset counts that would otherwise drift with the source.

What's changed

• update readme with v1.1 features and hardware coverage #36
• add user flow diagram and soften preset counts in readme #37

Full Changelog: v1.1.0...v1.1.1

v1.1.0 — polish round

Polish round on top of v1.0.0. Adds 13 new model presets (Llama 4 Scout/Maverick, Qwen 3, Gemma 3, Phi-4, DeepSeek V3 and more) and broadens hardware coverage beyond Apple Silicon to NVIDIA GeForce/Datacenter, AMD Radeon and Copilot+ PC NPUs (30 presets total). New Hugging Face URL import that auto-fills params and architecture from safetensors metadata, with a config.json paste fallback for gated repos and share links that round-trip via ?hf=. EN/ES i18n, manual dark/light theme toggle with FOUC-safe persistence, shareable URL state, markdown export, comparison chart against curated models. Open Graph metadata, README infographic, GitHub Pages auto-deploy on push to main, PR and issue templates, branching strategy and trunk-based workflow documented. Tests: 74 unit + 41 e2e on every push. Lighthouse: 100/100/100/100 mobile, 94/100/100/100 desktop.

What's changed

Models and hardware

• add 13 new model presets (llama 4, deepseek v3, qwen 3, gemma 3, phi-4) #9
• add nvidia, amd and copilot+ pc hardware presets with grouped dropdown #10

Hugging Face import

• import any model architecture from a hugging face url #32
• let users paste a config.json to import gated hugging face models #31

UX and i18n

• add shareable url state via querystring and copy share link button #11
• add comparison chart showing what fits on current hardware #12
• add copy and download as markdown for current results #13
• add english spanish toggle for all ui text #14
• promote live demo to header badge and callout #18
• add author credit in footer and readme #19
• add open graph and twitter card metadata #22
• convert og image to png and tune title and description length #23
• add tops vs bandwidth infographic #24
• add manual dark/light theme toggle #25
• fix lighthouse a11y findings and document the audit #34

Infra, tooling and docs

• auto-deploy github pages on push to main #16
• fix repo url case in readme and footer #17
• add pull request template #20
• add bug report and feature request issue templates #21
• document wsl workflow for regenerating images on windows #27
• document branching strategy and branch protection in contributing #28
• wrap render-svg browser lifecycle in try finally #29
• add mermaid diagram of render-svg script lifecycle #30

Full Changelog: v1.0.0...v1.1.0

v1.0.0 — first stable release

Initial release of the LLM TOPS Calculator. Interactive calculator that estimates how many TOPS you need to run local LLMs (Llama, Mistral, Qwen, Gemma, Phi) at different quantizations on Apple Silicon. Includes 17 model presets, 14 Apple chip presets, unit tests (37) + Playwright e2e (12), GitHub Actions CI, html report artifacts on failure.