Kapwing vs Lucide
Side-by-side trajectory, velocity, and editorial themes.
Kapwing has bet the product on generative AI workflows and is now consolidating after retreating from its ethical-AI side project.
Kapwing has fully reframed itself around generative AI: Kai (the in-product AI assistant), MiniMax video model integration, and a steady drumbeat of added image and video models. The cadence of actual product releases has slowed in recent months; the published surface has shifted toward research posts (AI slop on YouTube), engineering culture (100% AI-coding-agent adoption), and post-mortems on side projects. The January 2026 shutdown of Tess.Design — their artist-royalty AI marketplace experiment — closes off the ethical-AI-marketplace branch and focuses the company on the core editor.
The trajectory is consolidation, not expansion. Tess being wound down is a strategic retreat; the company appears to have decided that competing on AI-art ethics is not where it wins. The video-editor-as-AI-canvas thesis (Kai + integrated model marketplace) remains the bet, and partnerships with model providers (MiniMax most recently) suggest Kapwing wants to be the front-end aggregator rather than train its own models.
Expect more model partnerships (likely an integration with one of the new video model releases) and continued investment in Kai as the orchestration layer. The slower release cadence on the changelog suggests core editor work is happening but isn't being announced — likely a Kai-driven feature consolidation rather than new shipping surfaces.
Lucide is in a steady icon-addition cadence — eight minor releases in six weeks, mostly community PRs.
Lucide is releasing roughly twice a week along the 1.x line, with each minor adding 1–6 community-contributed icons (blender, broccoli, sticky note variants, repeat-off, waves-vertical, folder-bookmark, astroid, heart-x, layers-minus, bell-check) and refining existing ones (text-cursor, landmark, candy-cane, volleyball). Framework adapters (Svelte, Vue, Angular, React) receive small fixes alongside. There is no structural work in this window — it is a contributor-flow optimization.
The project sits in healthy steady-state. Low-friction PR throughput, regular minor versions, and no large refactors suggest Lucide has settled into being the default fork-and-extend icon library for designers and the maintainers are protecting that contributor pipeline rather than pushing a roadmap. Framework adapter parity is being maintained in lockstep with the core icon set.
Expect another 1–2 icons (or refinement PRs) per minor over the next few weeks, plus a framework-adapter patch as upstream Svelte/Vite/Vue dependencies shift. No 2.0 cut is signaled in this window.
See more alternatives to Kapwing →
See more alternatives to Lucide →