Real-Time Video Gen Research Scientist — Equity Options

Cantina; United States; Remote; USD 150,000 - 230,000; Full time

Benefits offered by this job

Medical, dental, and vision

Company equity

42 days PTO

Parental leave

401(k) retirement plan

Lifestyle spending account

Lunch and snacks in office

One Medical membership

Job summary

Cantina Labs is a social AI company building real-time models that animate characters and enable expressive, realistic interactions. We seek researchers who will work on streaming generation, training, distillation, and optimization to keep models fast and reliable during interactive use.

You’ll have freedom to propose directions and turn ideas into concrete experiments. Join us to help create a future where AI-enabled characters respond naturally with nuanced facial expression, motion, and

Qualifications

• Experience training or fine-tuning video or related generative models.
• Ability to learn unfamiliar methods quickly and turn ideas into useful experiments.
• Good experimental judgment, ownership, and clear communication.

Responsibilities

• Develop real-time video generation models with emphasis on latency, continuity, and quality.
• Advance expression and motion by integrating appearance, gestures, and speech.
• Fine-tune and adapt models to new data and capabilities; evaluate objectives.
• Explore distillation, inference optimization, and data-driven improvements.
• Collaborate with researchers and share early results and demos.

Skills

Video model training

Experimentation

Clear communication

Job description

Cantina Labs is a social AI company building real-time models that animate characters and enable expressive, realistic interactions. We seek researchers who will work on streaming generation, training, distillation, and optimization to keep models fast and reliable during interactive use.

You’ll have freedom to propose directions and turn ideas into concrete experiments. Join us to help create a future where AI-enabled characters respond naturally with nuanced facial expression, motion, and