AI21 Labs

LLMs struggle with complex tasks. AI21 Labs' Maestro - the world's first AI Planning & Orchestration System - doesn't. Check out the below benchmarks to see how Maestro improves outcome accuracy. Sign up for the waitlist and learn more here: https://lnkd.in/e8HAScza #EnterpriseAI #AI21Maestro #AIPlanning #AIOrchestration #HumanX #TrustworthyAI

?? Meet Maestro: The First AI Planning and Orchestration System ?? Today, AI21 Labs announced Maestro - an AI system designed to deliver enterprise-grade AI that organizations can actually trust. Unlike traditional AI approaches that rely on “prompt-and-pray” methods or rigid workflows, Maestro plans, executes and validates. ?? Reliable: Scales inference-time compute, selects the best models and tools and rigorously verifies outputs - ensuring accuracy within defined latency and cost limits. ?? Faster Deployment: Automates AI solution development. Define requirements, connect tools, set a budget. Maestro handles the rest. ?? Adaptive: Learns each enterprise environment, runs offline simulations, predicts success rates and costs and finds the most effective execution strategy for each use case at runtime. ?? Transparent: Provides full visibility with execution traces and validation reports, ensuring clear, controllable and trustworthy AI-driven decisions.and validation reports, ensuring clear, controllable and trustworthy AI-driven decisions. ?? Proven Accuracy Gains. In IFEval, Maestro significantly boosts model accuracy: ? GPT-4o: 85% → 91.9% ? Claude Sonnet 3.5: 88% → 95.2% ? o3-mini: 92% → 95.7% For complex, multi-requirement tasks, Maestro increases accuracy by up to 50% and enables reasoning models like o3-mini to exceed 95% accuracy. ?? Learn more and join the waitlist for early access: https://lnkd.in/e8HAScza #EnterpriseAI #AI21Maestro #AIPlanning #AIOrchestration #HumanX #TrustworthyAI

We’ve been working toward this moment for months and now, it’s almost here! Tomorrow, Monday, March 10th, our co-founder and co-CEO Ori Goshen will be unveiling a major milestone for AI21 Labs’ at #HumanX and we can’t wait to share more with you soon!

We’ve been working toward this moment for months and now, it’s almost here! Tomorrow, Monday, March 10th, our co-founder and co-CEO Ori Goshen will be unveiling a major milestone for AI21 Labs’ at #HumanX and we can’t wait to share more with you soon!

?? Introducing AI21’s Jamba 1.6: The Best Open Model for Private Deployment Today we are thrilled to announce Jamba 1.6, our new open model family that outperforms Cohere, Mistral & Llama and rivals leading closed models. Jamba 1.6 sets a new benchmark for enterprise AI, delivering market leading quality, efficiency and security. ?? Outperforms Cohere, Mistral & Llama on the Arena Hard benchmark? ?? Fully private on-prem or VPC deployment ?? Lightning fast latency and unparalleled performance on long contexts ??? Market-leading 256K context window? ?? Model weights available on Hugging Face Enterprises no longer have to compromise on model quality, speed and security with AI21 Labs' Jamba 1.6. Read more here: https://lnkd.in/e5imfMNc #BuildOpen #AI21Jamba #JambaOpenModels #EnterpriseAI

?? AI Models Judging AI Models: What Could Possibly Go Wrong? ?? In our second episode of ‘Yet Another AI Podcast’ , AI21 Labs' Yuval Belfer sits down with Algorithm Developer, Noam Gat, to discuss whether AI models can effectively judge themselves and how we can leverage them to build better AI systems. ?? Who should watch? ML engineers, AI researchers and AI developers who want to understand how AI self-evaluation impacts LLM performance. Key Topics Discussed ?? LLMs as Judges > Understanding relative ranking, absolute ranking, and fine-grained evaluation ?? Why AI needs Judges > Best-of-N selection, revision flows, and data filtering ?? Benchmarking Judge Models > RewardBench, LMArena PPE, Nemotron & more ?? Training Reward Models > From DJPO to Bradley-Terry ranking ?? LLMs as Judges & Reasoning models > What we learned from benchmarking DeepSeek R1 Check out the full episode on YouTube linked in the comments below and subscribe for future episodes! #GenAI #LLMJudging #AI21 #RewardModels #AIevaluation #LLMs #AIResearch #MLengineering #AIpodcast

LLMs were never designed to plan. At AI21 Labs, we’re building the future of enterprise AI with a soon-to-be-announced system that goes beyond text generation - it thinks, plans and executes with precision. In this video, Or Dagan, AI21’s Chief Product & Strategy Officer, breaks down why enterprises need AI systems that are purpose-built for planning - from models to infrastructure to architecture - and how this shift will make AI more reliable, efficient and actionable. ??? Watch now and let us know in the comments how AI that plans will transform the way you work. #AI #EnterpriseAI #LLMs #AISystems

Counting down to HumanX 2025! Join AI21 Labs' co-founder and co-CEO Ori Goshen as he takes center stage for a major announcement you won’t want to miss. ?? ?? Monday, March 10 | 12:15 PM PT ?? Center Stage, HumanX, Las Vegas If you're attending, let us know in the comments below. See you there!

Yoav Shoham, is a Professor Emeritus of Computer Science at Stanford University and a pioneer in artificial intelligence with decades of experience shaping the field. He also happens to be our co-founder and co-CEO. To read about his thoughts on the future of AI check out his contribution to the Google Cloud report, ‘Future of AI: Perspectives for Startups 2025’ and tell us what you think?in the comments below. https://lnkd.in/exZEjdUc

The AI21 Labs team is heading to HumanX in Las Vegas (March 9-13)!??? Our co-founder and co-CEO Ori Goshen will be making a big announcement on center stage that you won’t want to miss.??? If you’ll be there, we'd love to connect. Schedule a meeting at the link in the comments or stop by Booth 309. Looking forward to seeing you there! #HumanX2025 #EnterpriseAI #AISystems