Skip to main content

ZTE's Nebula-GUI AI Assistant Achieves 90% Offline Accuracy

ZTE Revolutionizes Smartphone AI with Offline-Capable Assistant

When voice commands translate directly into completed tasks, smartphone interaction reaches new levels of efficiency. ZTE's Nebula-GUI artificial intelligence system, now deployed in flagship devices including the Nubia Z70 Ultra and Z80 Ultra, achieves this through deep OS integration that eliminates manual app navigation.

Image

Technical Breakthroughs in Edge AI

The 7-billion-parameter model recently scored 84.38 in authoritative GUI Agent evaluations, demonstrating particular strength in:

  • Offline operation: All processing occurs on-device
  • Complex task execution: Multi-step processes like transportation bookings
  • Speed: Response times surpassing cloud-dependent alternatives

Current implementation supports 30+ mainstream Chinese apps including 12306, Meituan, and WeChat, achieving 90%+ accuracy for common use cases.

Solving the Chinese GUI Data Challenge

ZTE engineers developed an innovative pipeline to address the shortage of quality training data:

  1. Automated screenshot collection
  2. Semantic annotation systems
  3. Synthetic instruction generation This framework created thousands of operational pathways while reducing production costs by 40% compared to manual methods.

The team then applied supervised fine-tuning (SFT) to transform a general multimodal model into a specialized GUI agent capable of:

  • Screen element recognition
  • Intent understanding
  • Dynamic error correction
  • Cross-app permission management

Commercial Deployment and Future Roadmap

The Nebula-GUI rollout represents a paradigm shift from reactive voice assistants to proactive digital agents. ZTE confirms expansion plans for:

  • E-commerce price comparison
  • Multi-modal travel planning
  • Cross-platform data aggregation These enhancements will further reduce manual intervention in daily smartphone use.

Key Points

  • 90%+ accuracy for offline task completion
  • Supports 30+ apps including major Chinese platforms
  • 7B parameter model outperforms larger cloud-based alternatives
  • Proprietary data generation system solves Chinese GUI scarcity
  • Next-gen features targeting complex multi-app workflows

Enjoyed this article?

Subscribe to our newsletter for the latest AI news, product reviews, and project recommendations delivered to your inbox weekly.

Weekly digestFree foreverUnsubscribe anytime

Related Articles

News

Baichuan Intelligence Powers Up Edge AI with Major Funding Boost

Chinese AI startup Baichuan Intelligence has secured hundreds of millions in fresh funding to accelerate its edge computing solutions. The company's MiniCPM models are already transforming smart devices from phones to cars through partnerships with industry giants like Huawei and Geely. This investment will fuel further innovation in bringing efficient, private AI directly to consumer devices.

December 23, 2025
Edge ComputingAI StartupsSmart Devices
News

Mythic Secures $125M to Power Next-Gen AI Chips That Could Outshine NVIDIA

California-based Mythic has landed $125 million in fresh funding to develop revolutionary analog AI chips promising 100x greater efficiency than traditional GPUs. Led by DCVC with backing from Honda and Lockheed Martin, the investment will accelerate Mythic's push into data centers, autonomous vehicles, and defense systems. The company's breakthrough 'Starlight' platform already enhances low-light imaging for military and robotics applications.

December 19, 2025
AI ChipsSemiconductorsEdge Computing
News

JD.com Steps Up AI Chip Game with Storage-Compute Integration

Chinese e-commerce giant JD.com is making strategic moves in edge AI chip development, with job postings revealing high-paying positions for in-memory computing chip design. These chips could power JD's robotics and smart home devices, signaling the company's serious push into hardware innovation. While JD remains tight-lipped, the lucrative salary packages—up to 100K monthly—speak volumes about their commitment to this emerging technology.

December 12, 2025
AI ChipsJD.comEdge Computing
SenseTime's NEO Breaks Multimodal Barriers with Leaner, Faster AI
News

SenseTime's NEO Breaks Multimodal Barriers with Leaner, Faster AI

SenseTime and NTU's S-Lab have unveiled NEO, a groundbreaking multimodal AI architecture that ditches traditional patchwork designs. This innovative approach slashes data needs by 90% while outperforming competitors on key benchmarks. By rethinking everything from attention mechanisms to position encoding, NEO delivers lightning-fast edge device performance under 80ms latency. The open-sourced models could redefine how we build compact AI systems.

December 3, 2025
AI InnovationMultimodal LearningEdge Computing
Tsinghua's Breakthrough: AI Models Shrink While Getting Smarter
News

Tsinghua's Breakthrough: AI Models Shrink While Getting Smarter

Researchers from Tsinghua University and Miracl AI have unveiled a game-changing way to measure AI efficiency. Their 'Capability Density' metric shows how smaller models can now outperform larger ones, with intelligence per parameter doubling every 3.5 months. This innovation is already powering faster responses in smartphones, cars, and smart homes while slashing energy use.

November 24, 2025
AI EfficiencyCapability DensityEdge Computing
Bright Light AI Shatters Records with 2 Million Downloads in Just 6 Days
News

Bright Light AI Shatters Records with 2 Million Downloads in Just 6 Days

Ant Group's new AI assistant Bright Light has taken the app world by storm, hitting 2 million downloads faster than any competitor. What makes users flock to it? The ability to create custom apps through simple conversation - no coding required. From homework helpers to car cost calculators, Bright Light is changing how we interact with AI.

November 24, 2025
AI AssistantApp DevelopmentAnt Group