A new, uncensored version of the Qwen 3.6-35B model, named Genesis APEX MTP, has been released. This model boasts impressive performance, handling up to 200k context without glitches and successfully managing complex, intersecting tasks. It offers safetensors support for Apple MLX conversion and recommends specific quantization methods like APEX and MTP-APEX for optimal use. AI
IMPACT Offers enhanced context handling and uncensored capabilities for users running local LLMs.
RANK_REASON Release of a specific, fine-tuned version of an existing open-source model. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →