Peter Griffin (Coding Expert) on Instagram: "China just dropped WAN 2.6 — an AI video model that generates full 15‑second cinematic videos from a single prompt with multiple scenes, consistent characters, proper transitions, and multi‑speaker dialogue with lip sync + matching sound effects. 🎥✨ The game‑changer? Reference video mode: upload a 2–10 second clip with audio, and it clones both the face and voice to use that same character consistently across future videos. No more random face‑swaps