Tencent Hy4 ties Grok 4.7 at 35 on AA Intelligence Index, ranks #2 from China
TencentHunyuan · x · 2026-11-05
Tencent's new Hunyuan Hy4 scores 35 on the Artificial Analysis Intelligence Index v5, tying Grok 4.7 (xhigh) and trailing only Kimi K3.1 Pro (46) among China-lab models.
- Agentic coding is where the gap closes fastest: 31% on FrontierCode 2.0 vs 32% for Claude Fable 5.1 (max), leading Chinese labs
- Reliability tax: an 89% hallucination rate on AA-Omniscience
- Pareto shift on cost: priced at $0.834/$2.501 per 1M input/output tokens ($0.78 per index task), a clear cost-capability advantage over comparable Western models
The analyst notes Tencent was previously seen as lagging but calls Hy4 "remarkable," crediting @ShunyuYao12 on the Hunyuan team.
Related event: Tencent Hunyuan Hy4 Scores 35 on AA Index, Second Among Chinese Labs(3 posts)→
More from coding & agent
- Eon launches Yukon serverless data engine, arguing AI adoption is bottlenecked by data infra, not models — EXM7777 · 2026-11-05
- Git Magic by Ben Lynn lands on ChapterPal with an AI tutor, in two editions — burkov · 2026-11-05
- OpenCV Live Demos a Robot 'Visual Cortex' with RealSense, Jetson, ROS 2 and VLMs — chrismatthieu · 2026-11-05
- BigCommerce API MCP Server lets AI assistants manage store products, customers, and orders — modelcontextprotocol · 2026-11-05
- EmailMCP monetizes AI email agents: 23 read tools free, 28 send tools at $9.99/mo — modelcontextprotocol · 2026-11-05
- Todoist's internal Doistbot does state-of-the-art code reviews for under $2 — dscape · 2026-11-05