adaptive-memory-multi-model-router 2.2.2 → 2.2.3
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +2 -0
- package/README_ja.md +36 -0
- package/README_zh.md +59 -0
- package/package.json +3 -3
package/README.md
CHANGED
package/README_ja.md
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
1
|
+
# A3M Router 🔀
|
|
2
|
+
|
|
3
|
+
**Intelligent LLMルーティング · 99.5%精度 · ML不要 · GPU不要 · 47プロバイダー**
|
|
4
|
+
|
|
5
|
+
マルチシグナル複雑度スコアを使用して99.5%のルーティング精度を実現し、機械学習なしで動作します。各クエリを最も安価な利用可能なモデルに自動ルートします。
|
|
6
|
+
|
|
7
|
+
## 対応中国LLM
|
|
8
|
+
|
|
9
|
+
| プロバイダー | モデル | 強み |
|
|
10
|
+
|------------|--------|------|
|
|
11
|
+
| **DeepSeek** | V3, Coder, Reasoner | コード+推論、オープンウェイト |
|
|
12
|
+
| **Kimi (Moonshot)** | Kimi-1.5 | 128Kコンテキスト、中国語最強 |
|
|
13
|
+
| **Zhipu AI (GLM)** | GLM-4, GLM-4V | 中国語+バイリンガル |
|
|
14
|
+
| **Qwen (Alibaba)** | Qwen2, Qwen2.5-Coder | 汎用+コード |
|
|
15
|
+
| **Yi (01.AI)** | Yi-1.5, 34B | バイリンガル+長コンテキスト |
|
|
16
|
+
| **MiniMax** | abab6.5 | 1Mコンテキスト |
|
|
17
|
+
|
|
18
|
+
## クイックスタート
|
|
19
|
+
|
|
20
|
+
```bash
|
|
21
|
+
npm install adaptive-memory-multi-model-router
|
|
22
|
+
npx a3m-router serve
|
|
23
|
+
```
|
|
24
|
+
|
|
25
|
+
OpenAI SDKを`http://localhost:8787/v1`に向けるだけで、コード変更不要。
|
|
26
|
+
|
|
27
|
+
## オープンソース
|
|
28
|
+
|
|
29
|
+
- **GitHub**: https://github.com/Das-rebel/adaptive-memory-multi-model-router
|
|
30
|
+
- **npm**: https://www.npmjs.com/package/adaptive-memory-multi-model-router
|
|
31
|
+
|
|
32
|
+
MITライセンス。
|
|
33
|
+
|
|
34
|
+
## タグ
|
|
35
|
+
|
|
36
|
+
`LLMルーティング` `オープンソースLLM` `マルチプロバイダーAI` `APIコスト最適化` `DeepSeek` `Kimi` `Qwen`
|
package/README_zh.md
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
1
|
+
# A3M Router 🔀
|
|
2
|
+
|
|
3
|
+
**智能LLM路由 · 99.5%准确率 · 零ML · 零GPU · 47家提供商**
|
|
4
|
+
|
|
5
|
+
开源LLM路由代理，使用多信号复杂度评分实现99.5%路由准确率，无需机器学习。自动将每个查询路由到最便宜的可用模型，支持47家提供商。
|
|
6
|
+
|
|
7
|
+
[English](./README.md) | [日本語](./README_ja.md)
|
|
8
|
+
|
|
9
|
+
## 核心指标
|
|
10
|
+
|
|
11
|
+
| 指标 | A3M Router | RouteLLM |
|
|
12
|
+
|------|:-----------:|:--------:|
|
|
13
|
+
| 路由准确率 | **99.5%** | ~85% |
|
|
14
|
+
| 包大小 | **19.5KB** | ~1.5GB |
|
|
15
|
+
| 启动时间 | **<100ms** | ~3s |
|
|
16
|
+
| GPU需求 | **无** | 需要 |
|
|
17
|
+
| 提供商数量 | **47** | 2 |
|
|
18
|
+
|
|
19
|
+
## 支持的中国LLM
|
|
20
|
+
|
|
21
|
+
| 提供商 | 模型 | 优势 |
|
|
22
|
+
|--------|------|------|
|
|
23
|
+
| **DeepSeek** | V3, Coder, Reasoner | 代码+推理，开源权重 |
|
|
24
|
+
| **Kimi (Moonshot)** | Kimi-1.5 | 128K上下文，中文最好 |
|
|
25
|
+
| **智谱AI (GLM)** | GLM-4, GLM-4V | 中文+双语 |
|
|
26
|
+
| **通义千问 (Qwen)** | Qwen2, Qwen2.5-Coder | 通用+代码 |
|
|
27
|
+
| **零一 (Yi)** | Yi-1.5, 34B | 双语+长上下文 |
|
|
28
|
+
| **MiniMax** | abab6.5 | 1M上下文 |
|
|
29
|
+
|
|
30
|
+
## 快速开始
|
|
31
|
+
|
|
32
|
+
```bash
|
|
33
|
+
npm install adaptive-memory-multi-model-router # Node.js
|
|
34
|
+
pip install a3m-router # Python
|
|
35
|
+
|
|
36
|
+
npx a3m-router serve # 启动代理
|
|
37
|
+
```
|
|
38
|
+
|
|
39
|
+
然后将任何OpenAI SDK指向 `http://localhost:8787/v1`，零代码修改。
|
|
40
|
+
|
|
41
|
+
## 成本对比
|
|
42
|
+
|
|
43
|
+
| 月查询量 | 纯GPT-4o | A3M Router | 节省 |
|
|
44
|
+
|:--------:|:---------:|:----------:|:----:|
|
|
45
|
+
| 10K | $34 | $12 | $22 |
|
|
46
|
+
| 100K | $341 | $124 | $218 |
|
|
47
|
+
| 1M | $3,411 | $1,236 | $2,175 |
|
|
48
|
+
|
|
49
|
+
## 开源地址
|
|
50
|
+
|
|
51
|
+
- **GitHub**: https://github.com/Das-rebel/adaptive-memory-multi-model-router
|
|
52
|
+
- **npm**: https://www.npmjs.com/package/adaptive-memory-multi-model-router
|
|
53
|
+
- **文档**: https://das-rebel.github.io/adaptive-memory-multi-model-router/
|
|
54
|
+
|
|
55
|
+
MIT许可证，无需账号，纯本地部署。
|
|
56
|
+
|
|
57
|
+
## 标签
|
|
58
|
+
|
|
59
|
+
`LLM路由` `开源LLM` `多提供商AI` `API成本优化` `DeepSeek` `Kimi` `Qwen` `智谱AI` `LLM网关` `AI代理`
|
package/package.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "adaptive-memory-multi-model-router",
|
|
3
|
-
"version": "2.2.
|
|
3
|
+
"version": "2.2.3",
|
|
4
4
|
"shortName": "A3M Router",
|
|
5
5
|
"displayName": "A3M Router - Adaptive Memory Multi-Model Router",
|
|
6
|
-
"description": "LLM router & AI gateway
|
|
6
|
+
"description": "LLM router & AI gateway with 99.5% routing accuracy — supports 47 providers including DeepSeek, Kimi (Moonshot), Qwen, Zhipu GLM, Yi, Baichuan, MiniMax, StepFun. Zero ML, 19.5KB. Multi-signal routing, semantic cache, guardrails, cost analytics. MIT. TypeScript SDK + Python SDK + OpenAI proxy.",
|
|
7
7
|
"main": "dist/index.js",
|
|
8
8
|
"bin": {
|
|
9
9
|
"a3m-router": "dist/cli.js",
|
|
@@ -146,7 +146,7 @@
|
|
|
146
146
|
"bugs": {
|
|
147
147
|
"url": "https://github.com/Das-rebel/adaptive-memory-multi-model-router/issues"
|
|
148
148
|
},
|
|
149
|
-
"homepage": "https://github.
|
|
149
|
+
"homepage": "https://das-rebel.github.io/adaptive-memory-multi-model-router/",
|
|
150
150
|
"scripts": {
|
|
151
151
|
"test": "node test.js && node test/provider-test.js",
|
|
152
152
|
"test:providers": "node test/provider-test.js",
|