Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

Story

Nace.AI Open-Sources Drex-Max v1 Multimodal Decision Model

AI summary

Nace.AI released the open-weight Drex-Max v1, a dense 27B multimodal model that scores fixed candidate answers instead of generating text or JSON. The model accepts text, JSON and high-resolution images, supports local deployment on NVIDIA GPUs and Apple Silicon M4 or M5 systems, and can use context lengths up to 262,144 tokens. Nace.AI reported roughly 400-millisecond latency, but the supplied reports contain no independent reproduction of its benchmark results.

In-depth insight

Drex-Max v1 provides a locally deployable alternative to generative models for routing, tool selection and other fixed-choice decision tasks, with reported support for both NVIDIA GPUs and Apple Silicon.

#Nace.AI#Drex-Max v1#Apple Silicon#NVIDIA GPU#llama.cpp#Ollama

Original · 2 min readRead the original
Group chat

No comments yet. Start the conversation.