Cohere vs TwelveLabs

Introducing North 2: AI without compromises

Cohere has announced North 2, its latest platform update with 15+ new features, positioned as enterprise AI without compromises.

North 2: Enterprise AI without compromises

Cohere's biggest upgrade yet — unifying enterprise-grade security, performance, and flexibility in one platform.

Aleph Alpha's Kolibri Lags Behind Open-Weight Leaders

Aleph Alpha has released Kolibri, a new open-source German-English AI model under the Apache-2.0 license, designed for mission-critical applications in public administration and industry. Kolibri features a Mixture-of-Experts architecture with 78 billion total parameters and about 3.46 billion active per token. Trained on 20 trillion tokens, with 23% in German and a German-optimized tokenizer, it supports context lengths up to one million tokens. The model excels in agentic workflows, RAG, and native tool calling, while emphasizing data sovereignty and compliance, including a blocklist of 4.5 million URLs and GDPR/EU AI Act adherence. Independent benchmarks, however, place it behind at least 20 other open-weight models. Aleph Alpha is merging with Cohere to form a transatlantic sovereign AI company, subject to regulatory approval. In a broader economic debate, Reid Hoffman argues AI infrastructure expansion is the 'only reason' the US avoids recession, though economists dispute this.

• Aleph Alpha released Kolibri under Apache-2.0, with 78B total parameters and ~3.46B active per token.
• Kolibri trained on 20T tokens, with 23% German data and a German-optimized tokenizer, supporting 1M context.
• Independent benchmarks rank Kolibri behind at least 20 other open-weight models.

TwelveLabs Marengo Now in Amazon Bedrock Knowledge Base

TwelveLabs, a video intelligence company, announced two major developments. First, its Marengo video model is now available as an embedding model in Amazon Bedrock Managed Knowledge Base, enabling managed semantic video search on AWS. This integration automates ingestion, indexing, and search, eliminating the need for custom infrastructure or vector databases, while supporting metadata filters and keeping data within the customer's AWS account. Second, TwelveLabs released Pegasus 1.6, a model for egocentric video understanding in physical AI applications, such as robots, drones, and autonomous vehicles. Pegasus 1.6 supports five workflows: action segmentation, dense caption labeling, quality scoring, search/curation, and consent/compliance flagging. It builds on Pegasus 1.5 with improved entity recognition and faster processing. The CEO, Jae Lee, emphasized the mission to help machines understand the world through video.

• TwelveLabs' Marengo model is now available as an embedding model in Amazon Bedrock Managed Knowledge Base for managed semantic video search.
• Pegasus 1.6, built on Pegasus 1.5, is released for physical AI, enabling egocentric video understanding for robots, drones, and autonomous vehicles.
• Pegasus 1.6 supports five core workflows: action segmentation, dense caption labeling, quality scoring, search/curation, and consent/compliance flagging.

Marengo 3.5: Precision & Control for Frontier Video Intelligence

New blog post announcing Marengo 3.5, a frontier video intelligence model, along with posts on Compliance by TwelveLabs and a partnership with Quickplay.

TwelveLabs Launches Compliance App for Video Screening

TwelveLabs, a video intelligence company, announced the general availability of Compliance by TwelveLabs, a new SaaS solution for media, entertainment, and broadcast teams to screen video content against regional and custom compliance standards. The tool uses AI to identify potential violations with supporting rationale, reducing review time from a week to minutes for a two-hour master. It integrates NVIDIA's Synthetic Video Detector for real-time media authenticity checks and targets false-positive rates of 15% or lower. The app debuts after TwelveLabs' $100 million Series B and is the first application built on its full-stack agentic video intelligence platform. Features include rule pack authoring, real-time reviewer workflows, synthetic media flagging, and API-first access.

• TwelveLabs announced general availability of Compliance by TwelveLabs, a video content compliance SaaS.
• The solution integrates NVIDIA's Synthetic Video Detector for real-time authenticity scoring.
• It reduces manual review time for a two-hour video from about a week to minutes.