SoftBank Launches AI-Powered Large Telecom Model for Network Automation

SoftBank has launched the Large Telecom Model (LTM), a domain-specific, AI-powered foundation model built to automate telecom network operations. From base station optimization to RAN performance enhancement, LTM enables real-time decision-making across large-scale mobile networks. Developed with NVIDIA and trained on SoftBankโ€™s operational data, the model supports rapid configuration, predictive insights, and integration with SoftBankโ€™s AITRAS orchestration platform. LTM marks a major step in SoftBankโ€™s AI-first strategy to build autonomous, scalable, and intelligent telecom infrastructure.
SoftBank Launches AI-Powered Large Telecom Model for Network Automation

SoftBank Launches Telecom-Centric Generative AI Foundation Model

On March 19, 2025, SoftBank Corp. announced the development of a new Large Telecom Model (LTM) โ€” a domain-specific generative AI foundation model built to enhance the design, management, and operation of cellular networks. Leveraging years of expertise and extensive network data, LTM serves as a foundational model for AI innovation across telecom operations.


Trained on a diverse set of datasets โ€” including internal operational data, expert network annotations, and management frameworks โ€” the LTM offers advanced inference capabilities tailored specifically for telecom environments. The model represents a major leap toward AI-native network operations, enabling automation, optimization, and predictive intelligence across the full lifecycle of cellular network management.

AI-Driven Base Station Optimization with LTM

To demonstrate LTM’s practical applications, SoftBank fine-tuned the model to develop AI agents for base station configuration. These agents were tasked with generating optimized configurations for base stations that were not included in the training data. The results were validated by in-house telecom experts and showed over 90% accuracy.

This approach drastically reduces the time needed for configuration tasks โ€” from days to minutes โ€” with similar or improved accuracy. Compared to manual or partially automated methods, the LTM-led models offer:

  • Significant time and cost savings
  • Reduction in human error
  • Scalability across thousands of network nodes

The fine-tuned LTM models are capable of supporting two primary use cases:

1. New Base Station Deployment

In dense urban areas like Tokyo, the model is used to generate optimal configurations for new base stations. It receives input such as the deployment location, existing nearby infrastructure, and network performance metrics, and outputs a set of recommended configurations tailored to maximize performance and coverage.

2. Existing Base Station Reconfiguration

In scenarios like large events that temporarily increase mobile traffic, the model is used to dynamically adjust base station settings. It recommends real-time configuration changes to handle the surge in demand and maintain quality of service.

LTM as the Foundation for โ€œAI for RANโ€ and Future AI Agents

LTM is not just a standalone modelโ€”it is also the foundational layer for SoftBankโ€™s broader โ€œAI for RANโ€ initiative, which focuses on using AI to enhance Radio Access Network (RAN) performance. Through continued fine-tuning, LTM will enable the creation of domain-specific AI agents capable of:

  • Automated network design
  • Adaptive resource allocation
  • Predictive maintenance
  • Performance optimization across the RAN

These AI agents are designed to be modular and context-aware, making them easier to deploy across different scenarios and geographies.

Collaboration with NVIDIA for LTM Performance Gains and Flexibility

To maximize LTMโ€™s performance, SoftBank partnered with NVIDIA. Training and optimization of LTM were carried out using the NVIDIA DGX SuperPOD, a high-performance AI infrastructure used for distributed model training.

In the inferencing phase, SoftBank adopted NVIDIA NIM (NVIDIA Inference Microservices), which yielded:

  • A 5x improvement in Time to First Token (TTFT)
  • A 5x increase in Tokens Per Second (TPS)

NVIDIA NIM also supports flexible deploymentโ€”whether on-premises or in the cloudโ€”offering SoftBank the agility needed for enterprise-scale rollouts.

SoftBank also plans to use NVIDIAโ€™s Aerial Omniverse Digital Twin (AODT) to simulate and validate configuration changes before theyโ€™re applied, adding another layer of safety and optimization to the process.

โ€œHuman AIโ€ Vision Realized Through LTM

The LTM is an embodiment of SoftBankโ€™s โ€œHuman AIโ€ concept, as proposed by its Research Institute of Advanced Technology (RIAT). This vision complements โ€œMachine AIโ€ and emphasizes the integration of human expertise with AI to streamline operations and decision-making in mobile networks.

LTM is designed not just as a model but as a knowledge system, reflecting the insights of SoftBankโ€™s top network specialists. By integrating LTM-based models with AITRASโ€”SoftBankโ€™s AI-RAN orchestratorโ€”the company aims to build a unified AI framework for operating virtualized RAN and AI systems on the same infrastructure.

AITRAS Integration and Future Roadmap

The orchestration layer, known as AITRAS, is central to SoftBankโ€™s strategy for converged AI and RAN operations. LTM-powered models will eventually feed into AITRAS, enabling intelligent orchestration of both virtualized and AI-native workloads on a unified platform.

This integration is a key part of SoftBankโ€™s plan to build autonomous and self-optimizing networks that can:

  • React to real-time events
  • Predict and mitigate performance issues
  • Continuously evolve based on AI-driven insights

As SoftBank continues development of AITRAS, LTM will serve as its cognitive engine, providing operational intelligence across all layers of the network.

Global Collaboration Fuels LTMโ€™s Telecom AI Expansion

The development of LTM was led by the SoftBank RIAT Silicon Valley Office in collaboration with its Japan-based R&D team. Looking ahead, SoftBank plans to strengthen its global partnerships to scale the adoption of LTM across international markets and contribute to the advancement of next-generation telecom networks.

SoftBank also envisions using LTM to enable new services, enhance operational agility, and deliver superior mobile experiences to its customers.

ย Industry Experts Weigh in on LTMโ€™s Impact in Telecom

Ryuji Wakikawa, Vice President and Head of the Research Institute of Advanced Technology at SoftBank, said: โ€œSoftBank’s AI platform model, the ‘Large Telecom Model’ (LTM), significantly transforms how we design, build, and operate communication networks. By fine-tuning LTM, we can create AI agents for specific tasks, improving wireless device performance and automating network operations. We will continue to drive innovation in AI to deliver higher-quality communication services.โ€

Chris Penrose, Vice President of Telecoms at NVIDIA, added: โ€œLarge Telecom Models are foundational to simplifying and accelerating network operations. SoftBankโ€™s rapid progress in building its LTM using NVIDIA technologies sets a strong example for how AI can redefine telecom operations globally.โ€

LTM Sets a New Standard for AI-Powered Telecom Infrastructure

With the introduction of its Large Telecom Model, SoftBank has laid the foundation for a next-generation, AI-powered telecom infrastructure. LTM not only enhances operational efficiency but also unlocks new possibilities for intelligent automation, predictive optimization, and scalable AI agent deployment.

As SoftBank continues to refine and expand this modelโ€”alongside its work on AITRAS and โ€œHuman AIโ€โ€”it is positioning itself as a leader in the future of AI-native mobile networks.


Recent Content

The collision of two digital titans – AI and Bitcoin are on a collision course. One optimises the future; the other burns through energy to preserve the past. As AI sharpens its tools – from tracing tainted coins to auto-generating smart contracts – it is exposing cryptoโ€™s inefficiencies and vulnerabilities. Bitcoin may not die, but AI could force it to evolve: or risk irrelevance in a world demanding speed, sustainability and real utility.
Singtel launches 5G+, introducing nationwide network slicing for both consumers and enterprises, a global first. This upgrade brings faster speeds, lower latency, stronger indoor coverage, and real-time cyber protection to over 1.5 million users. Singtel 5G+ enhances mobile connectivity with the 700MHz spectrum, priority plans, and app-based slicing for business-critical apps, aligning with Singaporeโ€™s Smart Nation goals.
ย Virgin Media O2 and Daisy Group have joined forces to form a ยฃ1.4B B2B telecom and IT services powerhouse, targeting UK enterprises with an integrated offering that includes private 5G, cloud, AI, and cybersecurity solutions. With Virgin Media O2 holding a 70% stake and Daisy 30%, the new entity aims to accelerate enterprise digital transformation, drive operational synergies, and compete against both traditional telcos and cloud-first players in a fast-evolving market.
OpenAIโ€™s Stargate projectโ€”a $500B plan to build global AI infrastructureโ€”is facing delays in the U.S. due to rising tariffs and economic uncertainty. While the first phase in Texas slows, OpenAI is shifting focus internationally with โ€œOpenAI for Countries,โ€ a new initiative to co-build sovereign AI data centers worldwide. Backed by Oracle and SoftBank, Stargate is designed to support massive AI workloads and reshape global compute power distribution.
Twelve major European telecom providers, including Vodafone and Deutsche Telekom, have jointly urged the EU to allocate the full upper 6GHz band (6.425โ€“7.125 GHz) for mobile use, citing the spectrumโ€™s critical role in future 6G deployment. With the U.S. and China already advancing in this area, operators warn that delays could jeopardize Europeโ€™s digital leadership and hinder next-generation connectivity infrastructure.
Dirty data in data centers undermines everything from AI accuracy to energy efficiency. With poor metadata, data drift, and dark data hoarding driving up costs and emissions, organizations must adopt DataOps, metadata tools, and a strong data culture to reverse the trend. Learn how clean data fuels smarter automation, compliance, and sustainability.
Whitepaper
As VoLTE becomes the standard for voice communication, its rapid deployment exposes telecom networks to new security risks, especially in roaming scenarios. SecurityGenโ€™s research uncovers key vulnerabilities like unauthorized access to IMS, SIP protocol threats, and lack of encryption. Learn how to strengthen VoLTE security with proactive measures such as...
Whitepaper
Dive into the comprehensive analysis of GTPu within 5G networks in our whitepaper, offering insights into its operational mechanics, strategic importance, and adaptation to the evolving landscape of cellular technologies....

It seems we can't find what you're looking for.

Download Magazine

With Subscription

Subscribe To Our Newsletter

Scroll to Top