TwelveLabs Pegasus 1.5 becomes available on Amazon Bedrock
TwelveLabs announced that its Pegasus 1.5 video language model is now offered through Amazon Bedrock. The model can process videos up to two hours long—double the previous one‑hour limit—and returns time‑based metadata as JSON, including start and end timestamps and user‑defined fields. It adds multimodal prompting with up to 30 reference images and on‑the‑fly processing from Amazon S3 or base64…
Key points
- Pegasus 1.5 processes videos up to two hours long, twice the previous limit.
- Returns time‑based metadata as JSON and supports up to 30 reference images.
- Available today on Amazon Bedrock in US West (Oregon); billing via AWS Marketplace.
Pegasus 1.5 supports English fully and partial support for a dozen other languages, analyzes objects, on‑screen text, and audio, and can handle video files up to 2 GB from S3 or 25 MB as base64. The service is billed through AWS Marketplace under TwelveLabs, appears on the AWS bill, and is currently available in the US West (Oregon) region. Documentation includes sample Python code, CloudWatch metrics for monitoring, and guidance on defining segment definitions for use cases such as news archive tagging, sports highlight detection, and streaming‑platform recommendations.
Pegasus 1.5 from TwelveLabs is now available in Amazon Bedrock
AWS Machine Learning Blog · 19 October 2026
Loading the full article…
This text was published by AWS Machine Learning Blog and written by Ron Tal. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Mercury 3 [Inception Labs] - Matching Luna/Haiku at 1000 tokens/sec · 2 src
- Kling AI releases version 4.0 with stereo audio and a lighter Flash variant · 1 src
- DeepSeek releases V4.1-Flash-1017 update with lower cost and higher DeepSWE score · 26 src
- doubleAI releases debut code execution model and raises $220 million · 2 src
- CERN and Mistral sign statement of intent to develop AI foundation model for physics · 1 src
Comments
via GitHub Discussions