Qwen3.8 Max catches Claude Opus 4.8 but Kimi K3 still scores higher for 25 percent less

Qwen 3.8-Max, Opus 5 show why scores miss cost | VentureBeat

Alibaba's Qwen 3.8-Max reveals token pricing doesn't predict reasoning model cost: time-budget gaps (5-16x) flip benchmark rankings, and high-effort escalation hits timeout (79% of failures) at 3.1x spend. For agentic routing, measure cost per completed task, not per-token price; more reasoning often wastes budget on timeouts.