Skip to main content

Intel Enables Baidu's ERNIE Bot 4.5 Edge Deployment

Intel Powers Baidu's ERNIE Bot 4.5 for Edge Computing

Baidu made waves in the AI community on June 30 by open-sourcing its ERNIE Bot 4.5 series, comprising ten cutting-edge models. This release includes:

  • MoE models with 47B and 3B activation parameters
  • A dense model with 0.3B parameters
  • Pre-trained weights and inference code for developer accessibility

Image

Technical Synergy with Intel

Intel demonstrated remarkable technical agility by completing edge-side adaptations through its OpenVINO toolkit within hours of Baidu's release. This optimization enables deployment on Intel's Core Ultra platform, leveraging:

  • Hardware acceleration
  • Cross-platform compatibility
  • Enhanced inference performance

The collaboration builds on a partnership dating to 2021, where teams adapted multiple PaddlePaddle models (including PaddleOCR and PaddleDetection) for OpenVINO deployment.

Performance Breakthroughs

The ERNIE Bot 4.5 series showcases significant advances in:

  • Multimodal understanding
  • Text generation
  • Logical reasoning

Independent benchmarks indicate the models outperform GPT4.5 in multiple categories while offering API costs at just 1% of GPT4.5's pricing.

Key Points

  • Baidu open-sourced 10 AI models with commercial-ready weights
  • Intel achieved same-day edge deployment via OpenVINO optimization
  • Models demonstrate superior performance to GPT4.5 at fraction of cost
  • Partnership enables direct PaddlePaddle-to-OpenVINO model conversion

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

Microsoft's Harrier Model Breaks Language Barriers with Open-Source Release
News

Microsoft's Harrier Model Breaks Language Barriers with Open-Source Release

Microsoft's Bing team has open-sourced its powerful Harrier multilingual embedding model, supporting over 100 languages. Trained on 2 billion examples plus GPT-5 synthetic data, this technology promises to revolutionize search and AI applications. Developers can now access three model sizes on Hugging Face, with integration planned for Bing and next-gen AI services.

April 8, 2026
MicrosoftNatural Language ProcessingOpen Source AI
Google's Gemma4 Goes Truly Open: What It Means for Developers
News

Google's Gemma4 Goes Truly Open: What It Means for Developers

Google has taken a bold step with its latest AI model Gemma4, adopting the Apache 2.0 license to give developers unprecedented freedom. This marks a significant shift from previous restrictive policies, allowing commercial use and modification without legal hurdles. The new model boasts improved performance and seamless integration with existing developer tools, potentially leveling the playing field for smaller companies in the AI race.

April 3, 2026
Gemma4Open Source AIGoogle
Tesla's AI6 Chip: A Game-Changer in Edge Computing
News

Tesla's AI6 Chip: A Game-Changer in Edge Computing

Elon Musk has revealed Tesla's next-gen AI6 chip, set to complete tape-out by December. This powerhouse promises performance matching dual AI5 chips while being optimized for Tesla's humanoid robots and self-driving taxis. With a $16.5B deal with Samsung for 2nm production, Tesla is betting big on hardware-software co-design. Musk also shared intriguing views on AI's future limitations shifting from chips to energy.

March 19, 2026
TeslaAI ChipsEdge Computing
Tongyi Lab's New AI Tool Brings Hollywood-Quality Dubbing to Everyone
News

Tongyi Lab's New AI Tool Brings Hollywood-Quality Dubbing to Everyone

Tongyi Lab has unveiled Fun-CineForge, a groundbreaking open-source tool that solves one of AI dubbing's toughest challenges - realistic multi-person dialogue. Unlike traditional text-to-speech models, this film-grade system syncs voices with lip movements, maintains consistent character voices, and delivers emotional depth. The secret lies in its innovative four-modality fusion architecture and high-quality CineDub dataset. Early tests show it outperforms existing solutions, marking a significant leap forward for video localization and content creation.

March 16, 2026
AI DubbingVoice TechnologyOpen Source AI
News

Lenovo's AI Tablet Leap: OpenClaw Goes Mobile

Lenovo shakes up the tablet market by bringing OpenClaw's AI capabilities to mobile devices. Their new lineup, including the Pro 13 and YOGA Pad Pro, features one-click local AI deployment - no cloud required. This move transforms tablets from entertainment gadgets into powerful productivity tools while keeping your data private. The tech giant promises more surprises at their March 18 launch event.

March 12, 2026
Edge ComputingAI TabletsLenovo Innovation
News

Mac Mini's Hidden Power: How Engineers Unlocked AI Training on Apple's M4 Chip

In a surprising breakthrough, engineers have cracked open Apple's Neural Engine capabilities, revealing that Mac Minis can do far more than just run apps. By reverse-engineering the M4 chip with Claude AI's help, researchers discovered these compact machines can efficiently train AI models - challenging the need for expensive GPU setups. The findings show energy efficiency up to 80 times better than professional-grade hardware, potentially democratizing AI development.

March 9, 2026
Apple SiliconAI HardwareMachine Learning