Snowflake Turns Model Routing Into a Data Governance Feature
shashib · x · 2026-08-20
Snowflake has introduced dynamic model routing within its Cortex AI Gateway, converting model selection from a developer choice into a platform-controlled data governance feature. Using an "advisor pattern," the system tasks a smaller model first and escalates to a larger one if needed, while a classifier predicts when to skip the small model based on history.
Internal tests show significant token efficiency gains: data pipeline agents used a third of the tokens compared to a frontier-model-only approach, and engineering teams closed 25% more pull requests with the same token volume. The feature is moving to private preview.
More from Infra
- SK hynix preps for post-HBM era with 3D stacking shift — zephyr_z9 · 2026-08-20
- Guangdong, Alibaba Sign Strategic Pact to Boost AI, Chips and Computing Power — pstAsiatech · 2026-08-20
- GPU Market Pricing Chaos: Spreads Wide Across Venues — AccBalanced · 2026-08-20
- Dual RTX 3090 running Qwen3.8-27B locally at 50-65 tok/s — is that normal? — sugarfreecaffeine · 2026-08-20
- AVX-512 Deemed Superior to ARM SVE: Register & Masking Edge — lemire · 2026-08-20
- Modular open sources platform repo integrating MAX engine and Mojo language — modular · 2026-08-20