Edge0 has released a preview of its Edge0-35B-A3B model, a 35 billion parameter sparse Mixture-of-Experts (MoE) large language model designed to run efficiently on devices with limited memory. The model requires under 3 GiB of active RAM by streaming experts on demand, enabling interactive speeds of 15 tokens per second. Instructions are provided for integrating this model with various libraries and applications, including MLX, LM Studio, and Hermes Agent, making it accessible for local use on devices like Raspberry Pi. AI
IMPACT Enables running advanced LLMs on edge devices and mobile phones, potentially democratizing AI capabilities.
RANK_REASON Frontier-lab model release with system card. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on Hugging Face Trending Models →
- Atomic Chat
- Edge0
- Edge0/Edge0-35B-A3B-preview
- Google Colab
- Hermes Agent
- Kaggle
- LM Studio
- Mlx
- mlx-lm
- OpenClaw
- Raspberry Pi
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →