A Reddit user shared their experience comparing DeepSeek-V4.1-Flash-1017-GSQ-RCO with Qwen 3.8-Flash-Next-GSQ-RCO, noting that DeepSeek completed a task in approximately 15 minutes. The user provided detailed hardware specifications, including four RTX 3080 GPUs with 20GB VRAM each, 128GB of DDR4 RAM, and a context window of 1M for DeepSeek versus 256K for Qwen. Performance metrics indicated a significantly higher tokens per second (tg/s) for Qwen in terms of decoding and processing, despite DeepSeek's larger context window. AI
IMPACT Provides a real-world performance comparison between two LLMs, offering insights for users choosing between them.
RANK_REASON User-generated benchmark comparison of two LLMs. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →