Shandong 'Qilu Qixing': seven vertical industry large models
Shandong unveiled seven vertical-industry large models as '垂域智能齐鲁七星' at a Jinan AI showcase on 2026-10-14 — 浪潮海若, 卡奥斯天智, 极视角星际, 云鼎, 星港, 尚智, 奇智孔明 — covering government, petrochemicals, mining, ports and metallurgy.
Why it matters
Seven provincial vertical LLMs — 浪潮海若, 卡奥斯天智, 极视角星际, 云鼎, 星港, 尚智, 奇智孔明 — debuted as a branded matrix ('齐鲁七星'), each anchored in a real industrial deployment (ports, mines, steel, petrochemicals) with published efficiency numbers.
What you could build with it
A systems integrator serving chemical plants or port operators could pair one of these vertical models with on-site sensor data and existing SCADA logs to build an equipment-failure or logistics-bottleneck assistant. The result: domain-tuned operations copilots for industries that frontier labs do not serve directly.
Does it hold up?
Evidence is deployment-based rather than open: real published numbers exist for 星际 (aluminum-plant defect detection 95%+; 青岛地铁 job time 12h→3h) and 天智 (延长石油, 38 agents, ~20% efficiency lift), but all cases are vendor- or media-reported, the models are closed, and access is enterprise sales only.
Built with Shandong 'Qilu Qixing'
- 成为'AI视觉大模型第一股'，极视角做对了什么？blog · Third-party feature on 极视角/星际 with concrete deployments: 50+ model-collaboration vision scenarios at an aluminum giant (defect detection raised from ~50-60% to 95%+), and as the visual backbone of 青岛地铁's agent fleet (供电-agent job time cut from 12h to 3h).
- 智赋工业・数兴齐鲁丨垂域大模型看山东：一个产业共识如何从车间走向会场blog · Third-party analysis of all seven models with real deployment numbers: 天智 built 38 industrial agents at 延长石油, 云鼎 closed a data-model-agent loop across 10 mining specialties, 尚智 put 105 of 215 listed scenarios into production at 山钢, and 奇智孔明 tops IDC's China industrial-LLM app market share.
- AI赋能新型工业化 推动制造业迈向智能化新阶段blog · 上海证券报 reporting on 卡奥斯's 天智 petrochemical vertical model at 延长石油: 38 industrial agents across oilfield extraction, refining and safety, with reported ~20% average efficiency gains and 60% first-pass gasoline-blending success improvement.
Learn more
First spotted on article: source.
More AI models
OpenAI publishes 722 AI-generated math manuscripts from an unreleased modelOpenAI released 722 mathematical manuscripts in 372 result families to GitHub under Apache-2.0, produced by an…models · JEV 0.74ElevenLabs Eleven v4 + v4 Turbo: new emotive TTS architecture with a 100ms real-time variantElevenLabs released Eleven v4 and Eleven v4 Turbo on September 28: a new text-to-speech architecture with inline…models · JEV 0.72Mistral Large 4 (Le Chonk): 1.05T-param open-weight MoEMistral's new flagship: a 1.05T-parameter mixture-of-experts multimodal model with 49B active params and 1M-token…models · JEV 0.72VoiceStudio: open-source, fully-local ElevenLabs alternative with voice cloning, dubbing and dictation in 646 languagesVoiceStudio is an open-source, fully-local voice AI studio: voice cloning from a 3-second sample, voice design, video…models · JEV 0.71JetBrains Mellum 2.1: 12B MoE coding model trained for agentsJetBrains released Mellum 2.1, a 12B MoE model (2.5B active parameters, 128K context) under Apache 2.0, trained mainly…models · JEV 0.69DeepSeek V4-Flash official API: public beta with upgraded agent capabilities, Responses API and Codex supportDeepSeek's official V4-Flash is now a public beta API with massively upgraded agent capabilities - benchmark scores…models · JEV 0.69
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.