Files
nexus/wiki/concepts/Semantic-Routing.md
2026-05-03 05:42:12 +08:00

50 lines
1.5 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
---
title: "Semantic Routing"
type: concept
tags: []
sources: [engineering-autonomous-optimization-architect]
last_updated: 2026-05-01
---
# Semantic Routing
## Definition
语义路由——基于任务语义和历史性能数据,动态选择最优 LLM 提供商进行路由,而非使用固定规则或纯价格优先策略。
## Routing Decision Factors
- **任务类型**:文本提取 vs. 代码生成 vs. 创意写作 vs. 复杂推理
- **历史准确率**:该模型在此类任务上的 LLM-as-a-Judge 评分
- **成本**:每百万 Token 成本
- **延迟**P95 响应时间
- **可用性**:当前熔断器状态
## Example
```typescript
// Gemini Flash 在特定提取任务上准确率达 98% 且成本仅为 Claude Opus 的 10%
// → 自动将此类任务路由到 Gemini Flash
const rankedProviders = rankByHistoricalPerformance(taskType, providers);
for (const provider of rankedProviders) {
if (!provider.circuitBreakerTripped) {
return await provider.execute(task);
}
}
```
## vs. Traditional Routing
| 维度 | 固定路由 | 语义路由 |
|------|----------|----------|
| 成本优化 | ❌ | ✅ |
| 任务适配 | ❌ | ✅ |
| 动态适应 | ❌ | ✅ |
| 新模型集成 | 需人工修改 | 自动发现 |
## Related
- [[Autonomous-Optimization-Architect]]:实施语义路由的核心 Agent
- [[AI-FinOps]]:语义路由是成本优化的主要手段
- [[Shadow-Traffic]]:语义路由的评分数据来源于影子测试