Qwen3.8, an open-weight model, has demonstrated performance comparable to proprietary models like Claude Opus. While Claude Opus scored 63.1 on a benchmark, Qwen3.8 achieved 57.7, a gap of 5.4 points. Furthermore, Qwen3.8 is significantly more cost-effective, costing four times less per million output tokens. AI
IMPACT Highlights the increasing competitiveness of open-weight models against leading proprietary systems in terms of both performance and cost-efficiency.
RANK_REASON Comparison of open-weight vs proprietary model performance and cost. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Mastodon — sigmoid.social →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
