Qwen3.8 Max now ranked as the best overall model by agentic index

	Qwen3.8 Max now ranked as the best overall model by agentic index(artificialanalysis.ai)
	533 points by apitman 15 days ago \| 341 comments
	tl;dr: Summary not available.
	HN Discussion: ↑China has caught up to SOTA and Qwen's smaller local models are promising ↓The ranking/benchmark itself appears inconsistent or unreliable based on direct observation ↑Qwen performs impressively well in real-world troubleshooting and agentic tasks ~Any benchmark ranking Opus 5 highly loses credibility given poor real-world performance ↓Qwen is actually sloppy and unreliable in practice, contradicting the ranking