Hugging Face's logo
Activity Feed

AI & ML interests

XingChen-AGI is the open-source organization, dedicated to building agent-era large language models. Our flagship Xing4.0-29B-A4B is a 29B MoE model with only 4B activated parameters, natively supporting 256K context (extensible to 512K). With 4-bit quantization, it runs on a single consumer GPU (~15GB VRAM) and scores 93.52 on SuperCLUE Agent benchmark.

Recent Activity
summergrove  updated a dataset about 8 hours ago
XingChen-AGI/Xing4.0-29B-Intent-oss
summergrove  updated a model about 8 hours ago
XingChen-AGI/Xing4.0-29B-Intent-GGUF
summergrove  updated a model about 8 hours ago
XingChen-AGI/Xing4.0-29B-Intent-FP8
View all activity

XingChen-AGI 's collections 3

Xing4.0-29B-Intent
A 29B MoE with 4B activated that beats Gemini-3.6-Flash and GPT-5.6-Sol on intent recognition and extraction, plus a million-scale open dataset.
Xing4.0-29B-A4B
Native 256K context (expandable to 512K). 4-bit quant ~15GB VRAM, local run on RTX3090/4090. Scores 75.0 on SWE-bench Verified, 93.52 on SuperCLUE.
Xing4.0-29B-Intent
A 29B MoE with 4B activated that beats Gemini-3.6-Flash and GPT-5.6-Sol on intent recognition and extraction, plus a million-scale open dataset.
Xing4.0-29B-A4B
Native 256K context (expandable to 512K). 4-bit quant ~15GB VRAM, local run on RTX3090/4090. Scores 75.0 on SWE-bench Verified, 93.52 on SuperCLUE.