Exclusive
How Amazon weaned Alexa off Anthropic's pricey models to slash AI costs
By
Eugene Kim
You're currently following this author!
How Amazon weaned Alexa off Anthropic's pricey models to slash AI costs
Internal documents show Amazon worked to lower Alexa+ AI costs by reducing the use of Anthropic models and getting more work from every GPU.
TL;DRAI
Amazon redesigned Alexa to reduce Claude reliance, routing requests to in-house models and caching—AWS costs hit $1.7B in 2026, triple prior year. Inference efficiency now drives competition over model capability; multi-model routing becomes standard architecture.
1,351 words~6 min read
