A Smart Routing and Unified Adaptation Method for Large Language Models

By combining declarative configuration and a strategy engine, zero-code access and intelligent routing of large language models are achieved, solving the problems of heterogeneity and high integration costs when accessing and managing multiple models, improving the maintainability and scalability of the system, and ensuring call stability.

CN121279400BActive Publication Date: 2026-03-06NANJING INFORMATION HIGH-SPEED RAILWAY RES INST OF SCI AND TECH
2 Cites 0 Cited by

Patent Information

Application Number
CN202511842096.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-12-09
Publication Date
2026-03-06
Estimated Expiration
2045-12-09

AI Technical Summary

Technical Problem

Existing technologies suffer from interface heterogeneity, capability and pattern differences, high integration costs, poor scalability, and lack of governance capabilities when accessing and managing multiple large language models, making it difficult to achieve zero-code access and dynamic adaptation.

Method used

It adopts a configuration and strategy engine-based approach, defines model access details through declarative configuration files, provides a unified API interface, and combines strategy engine and JSON path technology to achieve zero-code access and intelligent routing for large language models. It supports dynamic configuration and hot updates, and integrates enterprise-level governance capabilities.

Benefits of technology

It enables zero-code integration of large language models, shields heterogeneity, provides a unified request and response abstraction layer, dynamically selects the optimal model, improves system maintainability and scalability, and ensures call stability.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This invention discloses a method for intelligent routing and unified adaptation of large language models. It defines all access details of the model through a declarative configuration file, allowing new models to be added without writing any code, achieving zero-code access for large language models. It provides a completely consistent calling interface for upstream applications, shielding the heterogeneity of all downstream large language models and constructing a unified request and response abstraction layer. Through a strategy engine and JSON path technology, it accurately handles complex streaming responses, including thought content, with intelligent parsing and content extraction. It supports dynamic configuration and intelligent strategy hot updates. It achieves intelligent model routing, dynamically selecting the optimal large language model instance. Simultaneously, it provides enterprise-level governance capabilities, integrating circuit breaking, degradation, rate limiting, and monitoring functions to ensure stability for model calls. This comprehensively improves system maintainability, scalability, and user experience consistency.
Need to check novelty before this filing date? Find Prior Art