AI BriefingAI21 Labs news
The latest AI and developer news from AI21 Labs, with short summaries.
Visit website →AI21 Achieves Top Score in Stanford University's Transparency Index
AI21 Labs recorded the highest score among commercial LLMs in Stanford University's Foundation Model Transparency Index (FMTI).
AI21's Jamba-Instruct Integrated into Snowflake Cortex
AI21 has integrated its latest foundation model, **Jamba-Instruct**, into **Snowflake Cortex** to support enterprise GenAI adoption.
Contextual Answers Outperforms Foundation Models in Question-Answering Performance
AI21 Labs' Contextual Answers model surpassed leading Foundation Models, recording superior results in question-answering performance and reliability metrics.
Jamba-Instruct Launches on Amazon Bedrock
AI21 Labs' latest model **Jamba-Instruct** has been added to **Amazon Bedrock**, supporting efficient large-scale context processing.
AI21 Begins Offering Jamba-Instruct on Microsoft Azure AI
AI21's latest foundation model **Jamba-Instruct** has launched via Microsoft Azure AI's **Models-as-a-Service (MaaS)**.
Is the Transformer Era Coming to an End?
AI21 Labs' Jamba model addresses the long-context processing limitations of Transformers through a hybrid architecture.
Reinventing the Future of AI with Long Context Jamba Instruct
AI21 hosted a hackathon leveraging Jamba's 256K context window, showcasing innovative AI use cases.
Long Context
AI21 Labs announced the Jamba 1.5 models, which maximize effective context length and improve cost efficiency.
Conversational AI Powered by RAG
This introduces a method for building conversational RAG that combines multi-turn dialogue with dynamic planning to enhance the reliability of enterprise GenAI.
Jamba-Instruct Now Available in Snowflake Cortex AI
Snowflake has integrated AI21's Jamba-Instruct, which supports a 256K context window, into Cortex AI.
7 Trends in Enterprise Long Context
Long context technology is drawing attention as a key factor determining the scalability and accuracy of enterprise AI.
Why Companies Struggle with AI Integration
The core barrier to AI integration isn't cost or hallucination, but a knowledge gap and a lack of product creativity.
JambaCare
AI21 unveiled JambaCare, which compares and analyzes insurance plans using a 256k context window.
AI21 Unveils Jamba 1.5 Model Family
AI21 has unveiled the **Jamba 1.5** family, a high-performance open model lineup based on an SSM-Transformer architecture.
Annual SOC 2 Audit Report
AI21 Labs has successfully completed its annual SOC 2 audit to comply with data protection and security standards.
vLLM CUDA Integer Overflow
A 32-bit overflow in vLLM's Mamba CUDA kernel caused logprob mismatches.
Enterprise AI Deployment
Enterprise AI is shaped less by model performance and more by system design, evaluation, and data preparation.
Orchestrating Modular AI Agents
A modularized agent architecture enables stage-by-stage evaluation and error tracing.
Padding-Minimizing Efficiency
AI21 reduced padding overhead by about 90% without changing the model.
Scaling vLLM Without OOM
Tuning and autoscaling vLLM for bursty traffic reduced throughput bottlenecks and latency.