Introducing Drex 2.0

Introducing Drex 2.0 — our fastest and smallest decision model!
Our first 35B-A3B MoE Mixture-of-Experts model, built for fast, accurate decisions.

Drex 2.0 is now available for dedicated deployment with @FireworksAI_HQ’s FireAttention!
• Up to 38% faster on 32K–128K inputs vs. Drex 1.5
• $0.06 per 1M input tokens

Try it out in Drex/Decide section our in your code https://nace.ai/developers.