Lingyun Smart Mining completed a Pre-A round of financing worth tens of millions of dollars, achieving the technical verification and commercial closed-loop of AI4Earth.
Recently, LynAI Mines Ltd announced the completion of a Pre-A round of financing worth tens of millions of US dollars. The investment was jointly made by leading financial investment institutions and well-known entrepreneurs, and existing shareholders continued to increase their investment. Light Source Capital has been a long - term and steadfast companion to the company's growth as an incubator. So far, LynAI Mines has completed three rounds of financing worth hundreds of millions of yuan in a year, fully demonstrating the capital market's high recognition of the feasibility of its technical route and the effectiveness of its commercialization progress.
This round of financing marks LynAI Mines' transition from the technology exploration stage to the large - scale verification and commercial implementation stage: Centered around the AI4Earth Intelligent Exploration Platform, the company has completed a complete commercial closed - loop from AI mineralization prediction to actual gold production on its own mining rights, and has officially established a replicable and verifiable new paradigm for intelligent mineral discovery.
This round of financing will be mainly used for expanding the pipeline of mining right projects in global core metallogenic belts, accelerating the iteration, verification, and deployment of multi - mineral models on the AI4Earth platform, and promoting the large - scale mining of in - production projects.
AI4Earth: An Emerging Industry Proposition
In the past decade, AI for Science (AI4S) has gradually evolved from an academic discussion to an industry consensus, profoundly changing the fundamental methodologies in fields such as protein structure prediction, new drug research and development, and material design. AI4S has proven one thing: when artificial intelligence truly penetrates the core workflow of a knowledge - intensive field, it brings not just a partial improvement in efficiency, but an overall shift in the discovery paradigm.
Today, a similar shift is occurring in the mining industry. LynAI Mines refers to this direction as AI for Earth (AI4Earth) — upgrading AI from an analytical tool in earth science to an engine for discovery, reasoning, and verification of complex earth systems. The mining industry has become the first entry point for the implementation of AI4Earth because mineral exploration and development activities naturally form the most dense, well - labeled, and economically verifiable data ecosystem in the earth system: from regional remote sensing, airborne magnetic and gravity surveys, geochemical anomalies, to boreholes, cores, grades, ore - body models, and production data, each layer of information records how deep - seated processes are transformed into shallow - seated resource expressions. AI4Earth is not just an indicator in papers, but a closed - loop implementation across the entire chain of mining right prediction, drilling verification, and mineral production.
The core judgment of AI4Earth is: The earth's mineral resources have never been truly exhausted; they have just exceeded the recognition boundaries of existing cognitive tools. Traditional exploration relies on the experience accumulation and linear inference of geological experts, and there are insurmountable processing bottlenecks when dealing with massive, heterogeneous, and multi - scale geological information. The continuous decline in the number of new discoveries of large - scale ore deposits globally is not because there are truly fewer resources, but because the methodology has reached its ceiling. What AI4Earth aims to do is to replace experience - driven with data - driven, and single - point judgment with systematic reasoning, fundamentally reconstructing the way of mineral discovery.
LynAI Mines is the earliest systematic proponent and practitioner of the AI4Earth concept. Around this concept, the company uses ultra - large - scale multi - modal geological data as the foundation, and specialized ore - forming prediction models trained for different deposit types as the core. Coupled with the on - site verification capabilities of unmanned aerial vehicle multi - sensor collection and ANT passive - source seismic imaging, it has built an end - to - end intelligent framework covering the entire chain of exploration decision - making. The key feature of this framework is its closed - loop nature: data drives the model, the model guides verification, and the verification results are fed back in real - time to strengthen the model, forming a self - iterating flywheel. Different from existing single - point tools in the market, the AI4Earth framework is not just a faster information processor, but a mineral discovery system with self - learning ability.
Project Achievements: Technical Verification and Commercial Closed - Loop
It is easy to propose a framework, but difficult to implement it in real projects. In the past year, LynAI Mines has continuously promoted the practical testing of the AI4Earth framework on multiple mining rights, accumulating experience in both technical verification and commercial closed - loop aspects.
The Toronto Gold Mine Project in the Mutare area of Manicaland Province, Zimbabwe, is the most complete end - to - end practice of the AI4Earth framework to date: From regional data scanning and AI target area selection, to ANT deep - structure confirmation, and then to the promotion of mining, the project has a resource volume of 1,760 kilograms. It is expected to produce gold for the first time in the second quarter of 2026 and recover the investment within a reasonable period. This is an end - to - end chain from data analysis to physical output of the AI4Earth framework, which proves that AI4Earth can not only find minerals more accurately, but also convert the judgment of mineral - finding into quantifiable economic returns.
Distribution map of gold grades in the target area: Sampling sections with Au grade > 0.2 g/t are continuously distributed along the contact zone (strike extension > 500 m, locally > 800 m)
Gold ore specimens and mineralized outcrops in Mutare Toronto
Mutare Toronto Gold Mine site (May 2026): Open - pit waste stripping and mining transport vehicle operations are in progress, and heap leaching and stope platform construction are underway
In Western Australia, LynAI Mines' Malcolm Gold Project officially entered the 3,300 - meter RC drilling verification stage in April 2026. Malcolm is located in the Yilgarn Craton, one of the oldest and most important gold - forming belts in the world, with high historical production grades and significant potential for deep and strike extensions. AI4Earth undertakes a complete reasoning chain: The system completed the establishment of the mine area cognitive framework within 72 hours. Through the extraction and classification of altered minerals, lithology prediction, and the generation of a mineralization probability heat map, it finally identified new potential target areas outside the known ore points, and further confirmed them through the deep - structure imaging of ANT passive seismic technology. Drilling is currently in progress, and the results will provide publicly available and auditable model verification data, which will be rigorously tested by drilling in a high - quality historical data environment.
Meanwhile, the greenfield mining right projects promoted by the company in Australia represent another direction of AI4Earth's capabilities — Completely relying on the ore - forming prediction model to delineate target areas and complete mining right applications on blank plots without historical drilling data. This is an extension of AI4Earth from auxiliary judgment to active discovery, and it is also a capability dimension that very few teams in the industry have truly achieved.
LynAI Mines has currently formed a pattern of parallel promotion of multiple projects, covering different minerals and exploration stages, which together constitute the double - line closed - loop of the AI4Earth framework — AI empowerment and efficiency improvement of existing assets, and AI - driven discovery of incremental minerals — and has officially established LynAI Mines' market position as a pioneer in AI4Earth.
Business Model: Deeply Linked with Mineral Discovery
LynAI Mines has chosen a business path that is naturally compatible with the AI4Earth framework.
For early - stage exploration projects, the company exchanges AI technical services for project rights. The quality of technical judgment directly affects the value of mining rights, so the technical contribution can be converted into long - term rights and interests sharing rather than one - time service fees. This model deeply links the platform's commercial interests with the results of mineral discovery. At the same time, the real data generated by each project is fed back synchronously, continuously strengthening the prediction ability of the AI4Earth framework. For mature mining companies, the company provides platform - based subscription services, embedding AI capabilities into their existing exploration workflows. The two models complement each other, jointly supporting the continuous iteration of the AI4Earth framework in the global project pipeline.
Currently, LynAI Mines' project pipeline has extended to major metallogenic belts in Africa, Oceania, Central Asia, and South America, and it has completed technical verification cooperation with multiple mining enterprises, accumulating a complete path from technical judgment to capital operation in the mining capital market.
Wang Xuance, the founder of LynAI Mines, said: "We proposed AI4Earth not just to describe what we are doing, but because we believe this is a real shift happening in the entire industry. The gold production in Zimbabwe is the first complete proof of this logic from start to finish, and projects like Malcolm are the next. What we want to establish is not just a faster tool, but a system that makes mineral discovery more predictable, verifiable, and replicable — a discovery flywheel that can continuously strengthen itself."
Huang Xinxin, the person - in - charge of the incubation business of Light Source Capital, said: "LynAI Mines is the earliest team to systematically accumulate in the AI4Earth direction and is also the practitioner who has gone the deepest in the commercial closed - loop. A team that promotes data accumulation, model iteration, and real - project verification simultaneously and has completed the complete chain from prediction to gold production on its own mining rights is extremely rare in this track. The first - mover advantage in the data - intensive AI track often has decisive significance — once data, models, and verification experience form a flywheel, the cost for latecomers to catch up will increase non - linearly. We believe that LynAI is at a critical window when this flywheel starts to turn."