Fontsource Inter
Self-host the Inter typeface with a single npm install and zero third-party font requests.
Repository Health
Technical Analysis
@fontsource/inter packages the open-source Inter typeface — one of the most widely used interface fonts on the web — as a self-hostable npm module. Instead of pulling render-blocking requests from Google Fonts, you install the package, import the generated CSS, and serve the bundled WOFF2/WOFF files straight from your own build output.
It ships pre-generated @font-face CSS for every weight (100-900) in both normal and italic styles, plus subset-specific stylesheets (latin, latin-ext, cyrillic, cyrillic-ext, greek, greek-ext, vietnamese) so you only load the glyphs your audience needs. A companion variable-font package covers the full weight axis in a single file for teams that prefer one request over many.
What You Get
- Pre-generated @font-face CSS for every weight (100-900) in both normal and italic styles
- WOFF2 and WOFF binary font files bundled directly in the package, no CDN or external request required
- Subset-specific CSS (latin, latin-ext, cyrillic, cyrillic-ext, greek, greek-ext, vietnamese) for selective glyph loading
- SCSS mixins and a metadata/unicode JSON pair for tooling that needs to select subsets or weights programmatically
Common Use Cases
- Removing the Google Fonts network request to improve Core Web Vitals (LCP/CLS) and page-load reliability
- Self-hosting fonts in privacy-conscious apps (GDPR compliance, strict CSP) that cannot call out to Google’s font CDN
- Bundling Inter into a design system or component library shipped as a pinned, versioned dependency instead of a CDN link tag
- Serving only the subsets a target locale needs, cutting shipped font-file weight versus a single monolithic font file
Under The Hood
Architecture The repository is a monorepo of self-contained font directories under fonts/{google,variable,icons,variable-icons,other}/<name>/, each holding generated @font-face CSS split by weight, style, and subset (e.g. fonts/google/inter/400.css, cyrillic-400.css), matching binary files/.woff2 and files/.woff assets, per-font metadata.json/unicode.json/package.json, and scss/ mixins for Sass consumers. Root-level scripts/ (metadata.ts, fontlist.ts, algolia.ts, check-duplicates.ts, gfm-metadata.ts) aggregate every font directory’s metadata into top-level metadata/fontsource.json, FONTLIST.json, and Algolia search records used by fontsource.org; the actual CSS/font-file generation and npm publishing happen through the separate @fontsource-utils/cli and @fontsource-utils/publish tooling, so this repo is primarily a store of generated artifacts plus aggregation scripts rather than application logic.
Tech Stack Aggregation scripts are TypeScript executed via Bun (bun.lock, .mise.toml pins the Bun toolchain), depending on @fontsource-utils/cli, @fontsource-utils/publish, google-font-metadata, algoliasearch, consola, pathe, and json-stringify-pretty-compact. The published @fontsource/inter package itself has zero runtime JavaScript dependencies: it is pure CSS plus WOFF2/WOFF binaries and a stub index.d.css.ts for editor tooling, consumed via plain CSS imports or Sass mixins.
Code Quality There are no hand-written unit tests over application logic since this is a generated-asset repo; correctness is instead enforced by fontsource create-verify --all (from @fontsource-utils/cli) wired as the test script, plus a check-duplicates.ts script guarding against duplicate metadata across 2,000+ font families. Because CSS and metadata are generated from Google’s font source rather than hand-authored, manual-edit error risk is concentrated in the shared generation tooling rather than spread across each font’s files.
API Design Installing @fontsource/inter and importing @fontsource/inter/400.css (or the bare @fontsource/inter for the 400-weight default) requires zero configuration to get a working @font-face rule. The <weight>[-italic].css / <subset>-<weight>[-italic].css naming convention is identical across all 2,000+ Fontsource packages, so familiarity transfers directly between fonts. The tradeoff is package footprint: every subset/weight/style combination ships as its own CSS+WOFF2/WOFF pair (419 files, ~4.3MB unpacked for Inter), so consumers must cherry-pick specific imports rather than rely on automatic tree-shaking.
Used by 163 apps in this directory
Unkey
Developer Tools
An open-source developer platform for API infrastructure — issue and verify API keys, enforce global rate limits, route traffic through a gateway, and get per-key analytics and audit logs.
Unsloth
AI Assistants · AI Development
Run and fine-tune LLMs, diffusion, audio and embedding models on your own hardware, from a native desktop app, a browser UI, or a Python library.
Valmis
AI Agents
A security-focused, self-hostable AI agent platform for production work — agents run in isolated containers that access credentials only through a host-side proxy, so the AI itself never sees API keys or plaintext credentials.
VideoHighlighter
Video Editors
Local AI desktop tool that scores, explains, and cuts the moments that matter in raw footage — scene, motion, audio, object, and action detection, fully offline via Ollama and Whisper.
VideoHighlighter
Video Editors
Local AI desktop tool that scores, explains, and cuts the moments that matter in raw footage — scene, motion, audio, object, and action detection, fully offline via Ollama and Whisper.
Voicebox
AI Development · Productivity · Voice AI
Clone voices, dictate anywhere, and give AI agents your voice — all locally.
VoiceStudio
Mcp · Music Audio · Voice AI
Open-source, fully local ElevenLabs alternative for voice cloning, voice design, video dubbing, dictation, transcription and audiobooks, with a local API and MCP server for agents.
VoiceStudio
Mcp · Music Audio · Voice AI
Open-source, fully local ElevenLabs alternative for voice cloning, voice design, video dubbing, dictation, transcription and audiobooks, with a local API and MCP server for agents.
VoiceStudio
Mcp · Music Audio · Voice AI
Open-source, fully local ElevenLabs alternative for voice cloning, voice design, video dubbing, dictation, transcription and audiobooks, with a local API and MCP server for agents.