Last week, @pika_labs launched the Pika 2.0 text and image-to-video generation model.
Features
- Excellent text alignment: precise matching of text content, WYSIWYG (What You See Is What You Get).
- Stunning visual effects: delivering professional-grade visual experiences.
- Scene Ingredients: supports uploading images, including people, places, objects, etc., making generated content more personalized and consistent.

Quick Start
In the Tutorial, several practical templates are also provided:
Template Testing
I made a video and uploaded two photos of Elon Musk and Claire Elise. I feel this kind of template is still a pretty smart approach.

Comparison with Sora
I ran the same prompt on both pika and sora.
A female gymnast in a red leotard performs multiple twisting somersaults on a competition-grade trampoline in a standard gymnasium. She executes rapid aerial spins while rotating forward, her body gracefully twisting like a corkscrew in mid-air. Filmed from the front view at regular speed for 5 seconds, capturing her complete sequence of rotations and twists. The lighting is bright and even, similar to television broadcast lighting at gymnastics competitions. The background shows typical gymnasium elements - blue exercise mats, scoring tables, and other standard competition equipment. The gymnast demonstrates perfect technique, her arms tucked tight and legs straight as she spins vertically and horizontally through each rotation, maintaining precise control at the peak of each bounce. The camera remains steady, positioned at eye level to capture the full height and dynamic spinning motion of the routine.
Sora
Pika 2.0
A peaceful scene of an elderly woman sitting on a comfortable floral-patterned sofa in her cozy living room, working intently on a sudoku puzzle. She's wearing a soft red cardigan sweater and delicate wire-rimmed reading glasses perched on her nose. The warm glow from a nearby table lamp creates gentle shadows and highlights her silver hair and concentrated expression as she carefully writes numbers with a pencil. A side table holds her reading glasses case and a half-completed crossword puzzle. The camera angle is at eye level, capturing her profile as she leans slightly forward in focus. Her gentle smile suggests she's enjoying the mental challenge. The overall composition has a warm, intimate feeling with soft lighting that creates a comfortable, homey atmosphere.
Sora
Pika 2.0