Pika WAN 3.0 Hands-on: Precise Lip Sync and 35% Lower API Costs
TomLikesRobots · x · 2026-08-25
Early testing of Pika Labs' newly released WAN 3.0 model highlights impressive lip sync capabilities, generated from a single image and audio reference with natural mouth articulation. The reviewer also notes that Pika API Club offers transparent pricing that is consistently 35% less expensive than competitors. Official features include 20 reference images, enhanced visual/audio realism, and 30-second generations.
More from Multimodal
- Hybrid live-action + AI short film for Shaq's game anniversary — ring_hyacinth · 2026-08-25
- World Labs Marble Now Available in Figma Weave to Turn Media into 3D Worlds — theworldlabs · 2026-08-25
- Goblin portrait generated by Gemini — Adventurous_Trip3230 · 2026-08-25
- Prompt example: Hyper-realistic Backrooms subway video generation — cocktailpeanut · 2026-08-25
- Higgsfield AI Relight simulates studio lighting in videos — petewoodbridge · 2026-08-25
- 8-Minute Animated Film Generated in Just 14 Minutes via @fal — OdinLovis · 2026-08-25