Huawei Cloud releases the new generation ModelArts Next model training and deployment platform

robot
Abstract generation in progress

Mars Finance News, June 5: Huawei Cloud officially launched its next-generation model training and inference platform, ModelArts Next. It provides four core capabilities: RL services, confidential inference, model routing, and a model matrix.

MaaS model routing supports three strategies: cost-priority, effectiveness-priority, and balanced mode. Enterprise-grade RLaaS services enable task creation in one minute, full-process visual observability, and safeguards for training–inference consistency.

Confidential inference leverages hardware-level trusted execution environments to ensure that in highly sensitive scenarios such as AI coding and financial risk control, the data processed by models is “only in and never out.”

The model matrix capability delivers Day0 deployment of leading mainstream SOTA models (such as DeepSeek, Kimi, Zhipu GLM, etc.), and, together with Huawei’s self-developed model Pangu, its capabilities cover a wide range of scenarios including programming and multimodal applications. (Wide-angle observation)

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned