FeiX7

• Vulkan backend outperforms ROCm on Strix Halo (gfx1151) — llama.cpp benchmark

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 81 comments

• Why use Quants other than Unsloth

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 41 comments

• Best OpenClaw Alternative

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 26 comments

• Best Local Model for Openclaw

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 199 comments

• Qwen Models are such good models?

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 27 comments

• Local Claude Code with Qwen3.5 27B

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 122 comments

• RTX 5070 Ti (new) vs RTX 3090 / 3090 Ti (used) for LLM inference + clustering

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 19 comments

• Creating Pi Extension with Pi and Qwen3.5 27B

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 1 comments

• Qwen3.5 9B for Pixel 9/10 Pro

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 4 comments

• Local Agents running in claude code/codex/opencode perform better?

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 4 comments

• ministral-3-3b is great model, give it a shot!

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 27 comments

• Problems with local Agentic Browsers

  Posted by FeiX7@reddit | LocalLLaMA | View on Reddit | 12 comments