Edge0 has released Edge0-145B, a 145-billion-parameter sparse mixture-of-experts model designed to run on high-memory laptops. The model requires approximately 29GB of peak memory, with a significant portion of its parameters not active during inference. Edge0 claims performance competitive with cloud-based models on various benchmarks, though these results are from internal testing and have not yet been independently verified. The model is available under an open-source license, allowing for research and commercial use of its outputs. AI
IMPACT This release pushes the boundaries of on-device AI, potentially enabling more powerful local applications and reducing reliance on cloud infrastructure.
RANK_REASON Open-source release of a large-scale (145B parameter) sparse mixture-of-experts model with claimed on-device capabilities. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- DeepSeek-V4-Flash-0731
- DeepSWE 1.1
- Edge0
- Edge0-120B
- Edge0-145B
- GitHub
- GPQA-Diamond
- Hugging Face
- MacBook Pro
- Qwen3.8 Flash-Next
- Terminal-Bench 2.1
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →