Improve batch embedding throughput for faster indexing

Author: PatrickSys

CHANGELOG.md

@@ -1,6 +1,12 @@
 # Changelog
 
 
+## [1.3.2] - 2026-01-16
+
+### Changed
+- **Embeddings**: Batch embedding now uses a single Transformers.js pipeline call per batch for higher throughput.
+- **Dependencies**: Bump `@modelcontextprotocol/sdk` to 1.25.2.
+
 ## [1.3.1] - 2026-01-05
 
 ### Fixed

package.json

@@ -1,6 +1,6 @@
 {
   "name": "codebase-context",
-  "version": "1.3.1",
+  "version": "1.3.2",
   "description": "MCP server that helps AI agents understand your codebase - patterns, libraries, architecture, monorepo support",
   "type": "module",
   "main": "./dist/lib.js",
@@ -87,7 +87,7 @@
   },
   "dependencies": {
     "@lancedb/lancedb": "^0.4.0",
-    "@modelcontextprotocol/sdk": "^1.25.1",
+    "@modelcontextprotocol/sdk": "^1.25.2",
     "@typescript-eslint/typescript-estree": "^7.0.0",
     "@xenova/transformers": "^2.17.0",
     "fuse.js": "^7.0.0",

pnpm-lock.yaml

@@ -75,9 +75,12 @@ export class TransformersEmbeddingProvider implements EmbeddingProvider {
 
     for (let i = 0; i < texts.length; i += batchSize) {
       const batch = texts.slice(i, i + batchSize);
-      const batchEmbeddings = await Promise.all(batch.map((text) => this.embed(text)));
 
-      embeddings.push(...batchEmbeddings);
+      const output = await this.pipeline(batch, {
+        pooling: 'mean',
+        normalize: true
+      });
+      embeddings.push(...(output.tolist() as number[][]));
 
       if (texts.length > 100 && (i + batchSize) % 100 === 0) {
         console.error(`Embedded ${Math.min(i + batchSize, texts.length)}/${texts.length} chunks`);