Open-source Agent World Model generates 1,000 SQL-backed executable environments for agentic RL with benchmark-winning results.
FEB 13, 2026|10 min read
SuffixDecoding at Production Scale with Arctic Inference and vLLM
SuffixDecoding now delivers 1.96x–3.12x end-to-end speedups in vLLM and Arctic Inference with major CPU optimizations for fast, production-ready LLM serving.
DEC 02, 2025|9 min read
Accelerating PyTorch Innovation at Scale: Snowflake at PyTorch Conference 2025
How Snowflake tackles four core AI challenges: scaling deep learning, training thousands of models, accelerating inference, and balancing multilingual performance.
NOV 19, 2025|7 min read
What’s Your Agent’s GPA? A Framework for Evaluating AI Agent Reliability
Learn how the Snowflake AI Research team's Agent GPA framework achieved 95% error detection and 86% localization accuracy to accurately measure and debug agent performance.
NOV 04, 2025|8 min read
Optimizing Query Execution in Cortex AISQL
How we made AISQL up to 8x faster and 70x cheaper through AI-aware planning, adaptive model cascading, and semantic join rewriting.
NOV 04, 2025|8 min read
Smarter, Faster and Snowflake-Native: Real-Time Text2SQL Behind Snowflake Intelligence
Discover Arctic-Text2SQL-R1.5, Snowflake's new, native Text-to-SQL model built to overcome LLM latency and deliver higher accuracy for real-time conversational analytics.
NOV 04, 2025|7 min read
Fast Reasoning on GPT-OSS with Speculative Decoding and Arctic Inference
Snowflake AI Research boosted GPT-OSS reasoning speed by 1.7–1.8x using Arctic Inference with speculative decoding, enabling faster agentic AI.
AUG 25, 2025|4 min read
Bridging the Gap Between LLMs and Real-World Challenges: Snowflake at ACL 2025
Snowflake AI Research shares 4 breakthrough papers accepted at ACL 2025, advancing model efficiency, document AI, text-to-SQL reliability, and LLM evaluation.
JUL 30, 2025|7 min read
Arctic Long Sequence Training (ALST): Scalable And Efficient Training For Multi-Million Token Sequences
Snowflake's ALST enables scalable training of long-context models with up to 15 million tokens using Hugging Face and DeepSpeed, all without custom modeling code.
JUN 24, 2025|10 min read
Inside Snowflake Intelligence: Five Pillars of Enterprise-Grade Agentic AI
Explore the underlying architecture, orchestration, and system-level optimizations behind Snowflake Intelligence, a production-grade agentic AI system built for enterprise reasoning.
JUN 03, 2025|13 min read
1
2
3
4
5
