Alibaba Cloud released Qwen3.8-Max via API with detailed benchmark results showing performance roughly comparable to Claude Opus 4.8, with mixed wins across key tests (e.g., TerminalBench-2.1: 86.6 vs 84.6, SWE Bench Pro: 67.7 vs 69.2). The main competitive advantage is pricing: $2–6 per million tokens vs Claude's higher rates, plus $0.25 per million for cached tokens. Model weights release planned for next week. Relevant for SMEs evaluating AI infrastructure costs and developer-focused businesses.