Pegasus 1.6
Video modelCard summarizedPegasus 1.6 is a generative model. It analyzes videos or images to generate contextually relevant text.
Where to call it
- TwelveLabs API
pegasus1.6
Good for
- Action labeling of first-person footage from wearable and teleoperated cameras
- Quality control of robotics and physical-AI footage against expected behavior
- Trajectory and next-action prediction in teleoperated video
Overview
Pegasus 1.6
Pegasus 1.6 is a generative model. It analyzes videos or images to generate contextually relevant text.
- Max resolution
- 5184×2160
- Max length
- 2 h analyzed per request
- Modes
- video-to-text, video Q&A
- Released
- 6 Oct 2026
Pegasus 1.6 is TwelveLabs' video-language model: it watches a video, or one to twenty still images, and returns free text or schema-shaped JSON, including timestamped segments. Over Pegasus 1.5 it adds first-person (egocentric) video, native image analysis, more consistent naming of people, objects and characters, richer metadata extraction, timestamped events inside each segment, and segmentation without an output token limit. It is a drop-in upgrade: only model_name changes. Launch post (6 Oct 2026); Pegasus 1.6 docs page; release notes
pegasus1.6- Typical price
- $1.75 / hour
- Max resolution
- 5184×2160 (input video)
- Max length
- 2 h analyzed per request (source up to 4 h; 1 h on the sync endpoint)
- Languages
- English (full); Arabic, Chinese, French, German, Italian, Japanese, Korean, Portuguese, Russian, Spanish, Thai, Vietnamese (partial)
- Weights
- Closed
- Released
- 6 Oct 2026
- video-to-text
- video Q&A
- video summarization
- video segmentation (timestamped JSON)
- in-segment events
- egocentric video understanding
- image analysis (1–20 images)
- structured JSON output