laya-decision-api
by Community
把本地 System-1 决策模型（Laya）包成 HTTP 服务：意图分类 / 工单分派 / LLM 路由 / 内容审核，离线零 token、不生成文字；Apple Silicon(MLX) 与 Linux(torch) 双后端；可导出 OpenAPI 3.0/3.1；Java/TS/Python 零依赖 SDK。
OSS
laya-decision-api
Added 4 Oct 2026
Overview
Wraps the local System-1 decision model Laya as an HTTP service for intent classification, ticket dispatch, LLM routing, and content moderation. Runs offline with zero token usage and no text generation. Supports Apple Silicon (MLX) and Linux (torch) backends, exports OpenAPI 3.0/3.1, and provides zero-dependency SDKs for Java, TypeScript, and Python.
Best for
Best for
Developers needing lightweight, offline decision routing for internal services
Use cases
- Classify intents for customer support tickets
- Route requests to appropriate LLM models
- Moderate content without sending data to external APIs
Notes
Wraps the local System-1 decision model Laya as an HTTP service for intent classification, ticket dispatch, LLM routing, and content moderation. Runs offline with zero token usage and no text generation. Supports Apple Silicon (MLX) and Linux (torch) backends, exports OpenAPI 3.0/3.1, and provides zero-dependency SDKs for Java, TypeScript, and Python.
1 stars on GitHub. Last updated 2026-10-01. Licensed Apache-2.0.
Use cases
- Classify intents for customer support tickets
- Route requests to appropriate LLM models
- Moderate content without sending data to external APIs
Pros
- Offline and zero token cost
- Dual backend support for MLX and torch
- OpenAPI export and SDKs for multiple languages
Cons
- Very early stage with only 1 star on GitHub
- Community-maintained with limited documentation
- Requires local model setup and management
Indexed from awesome-llmops and enriched against its public facts.
Pros
- Offline and zero token cost
- Dual backend support for MLX and torch
- OpenAPI export and SDKs for multiple languages
Cons
- Very early stage with only 1 star on GitHub
- Community-maintained with limited documentation
- Requires local model setup and management
Get the free Developer’s Field Guide
A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.
Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.