DOCSEARCHbyAlgolia
Submit Docs

Alibaba Cloud AI and Search Documentation

website·Indexed 19 days ago

https://help.aliyun.com/

Comprehensive technical documentation for Alibaba Cloud's AI and search ecosystem. Includes PAI (Platform for AI) for end-to-end model development, Model Studio (Bailian), and a suite of search services including OpenSearch (Vector, Conversational, and High-Performance editions), AI Search Open Platform, and CleverSee. Also covers specialized AI services such as OCR, Image Search, Machine Translation, and NLP processing tools.

Tokens
737.9K
Snippets
879
Records
8.5K
Agent score
95%

What's inside Alibaba Cloud AI & Search

  1. Overview of Alibaba Cloud Model Studio

    Alibaba Cloud Model Studio is a unified model service platform that provides access to the Qwen series and other major third-party LLMs. It supports multimodal capabilities (text, image, audio, and video) and allows developers to call models via official Qwen APIs or OpenAI-compatible APIs without managing underlying infrastructure.

  2. Overview of EAS (Elastic Algorithm Service)

    EAS (Elastic Algorithm Service) is a high-performance, scalable AI model deployment service within Platform For AI (PAI). It allows developers to deploy trained models as online inference services or AI web applications. Key features include support for heterogeneous resources, auto-scaling, one-click stress testing, canary releases, and real-time monitoring to ensure stability in high-concurrency scenarios while optimizing costs.

  4. Overview of OpenSearch Vector Search Edition

    OpenSearch Vector Search Edition is a distributed C++ vector search engine designed for high-concurrency retrieval and large-scale production workloads. It supports real-time incremental writes with second-level visibility and multi-shard index building. The architecture utilizes QRS workers and Searcher workers to manage data and retrieval. It supports data sources from MaxCompute, Saro, and APIs, and provides cluster scaling via a control system.

  5. Overview of Artificial Intelligence Recommendation (AIRec)

    Artificial Intelligence Recommendation (AIRec) is a personalized recommendation service that provides REST APIs for e-commerce products, content feeds, and news articles. It includes pre-built industry templates for E-commerce, Content, and News, and supports two primary service types: 'You May Also Like' and 'Related Recommendations'. The service is available in two editions: Industry Operation Edition and Algorithm Configuration Edition. SDKs are available for Java, Python, Go, and PHP.

  6. Overview of Artificial Intelligence Recommendation (AIRec)

    Artificial Intelligence Recommendation (AIRec) is a personalized recommendation service that addresses cold-start and relevance challenges. It combines collaborative filtering, real-time behavioral signals, and industry-specific models tailored for E-commerce, Content, and Live Streaming scenarios. The service utilizes deep learning CTR (Click-Through Rate) models and a recommendation pipeline architecture to provide real-time inference.

  7. Overview of PAI Data Science Workshop (DSW)

    DSW (Data Science Workshop) is a cloud-based AI development IDE that supports three primary development environments: Notebook, VSCode, and Terminal. It comes pre-installed with mainstream AI framework images such as PyTorch and TensorFlow, supports various heterogeneous computing resources, and allows mounting of OSS, NAS, and CPFS datasets for efficient development.

  8. Overview of Artificial Intelligence Recommendation (AIRec)

    Artificial Intelligence Recommendation (AIRec) is a personalized recommendation service by Alibaba Cloud designed for businesses and developers. It leverages big data and AI to provide personalized recommendations for items, content, and news without requiring the user to build a recommendation engine from scratch. Users provide the necessary data and call AIRec API operations to retrieve recommendations.

  9. Overview of Moonshot AI models in Model Studio

    Alibaba Cloud Model Studio provides access to Moonshot AI models:

    • Moonshot-Kimi-K2-Instruct: A trillion-parameter MoE model (32B activated) with strong coding and tool-calling capabilities.
    • kimi-k2-thinking: A general agentic reasoning model specialized in deep reasoning and multi-step tool integration for complex problems.

  10. Overview of PAI-Rec (Personalized Recommendation Platform)

    PAI-Rec is a platform providing deep customization capabilities for the entire recommendation system pipeline. It is designed for enterprise developers to independently build, develop, iterate, and operate their own recommendation systems.

  11. Overview of Alibaba Cloud Billing Types

    Alibaba Cloud provides three primary types of billing records:

    Bill TypePurposeIssuance TimeAccess Link
    Monthly BillView total amount due for the month; export PDF for reconciliation and settlement.3rd of next month, 12:00Monthly Bill Overview
    Consumption DetailsView detailed consumption per resource for cost analysis; export as CSV.3rd of next month, 12:00Bill Details
    Split-Bill DetailsSplit bills by department/project for internal cost allocation (not for settlement).4th of next month, 12:00Split-Bill Details

  12. Overview of Platform for AI (PAI)

    Platform for AI (PAI) is an end-to-end AI development platform from Alibaba Cloud. It supports the entire AI lifecycle, including data labeling, model development, training, and deployment. Key components include:

    • AI Compute: General and Lingjun resources.
    • DSW (Data Science Workshop): Interactive modeling using JupyterLab or WebIDE.
    • DLC (Deep Learning Containers): Distributed training for single-node or multi-node GPU setups.
    • EAS (Elastic Algorithm Service): Deployment of models as online inference services.
    • Model Gallery: Zero-code deployment and training for open-source LLMs.
    • Designer: Visual AI development with 100+ algorithms.