multimodal

4 articles tagged with this topic

Qwen 27B Runs on Just 18GB VRAM — Local LLM Bar Drops Again

Qwen's 27B multimodal needs only 18GB VRAM (Q4), runnable on consumer GPUs. Local LLM bar drops again, but MoE version still demands clusters.

1d ago2 min read

DeepSeekopen-source-LLMs

DeepSeek V4 Vision Weights Delayed — China's Open-Source AI Lead Is Shrinking

DeepSeek rewrote rules with open-sourced V3 and R1. Now V4 Vision weights are silent. Question: how long can China's open-source lead last?

3d ago2 min read

DeepSeekV4-Flash-Vision-Exp

DeepSeek can finally see: multimodal nears Opus-4.8, pricing unchanged

DeepSeek launches V4-Flash-Vision-Exp with multimodal Agent reportedly near Opus-4.8 at text pricing—first Chinese model at the multimodal frontier.

Aug 222 min read

QwenMuse Glimmer

Qwen Loses to Muse Glimmer in Image Recognition: Is the Moat Cracking?

LocalLLaMA test: Muse Glimmer 30B swept Qwen 3.8 27B on OCR and image-text tasks. Informal, but exposes Qwen's multimodal weakness.

Aug 152 min read