@mastra/memory 1.0.1 → 1.1.0-alpha.0

package/CHANGELOG.md

@@ -1,5 +1,21 @@
 # @mastra/memory
 
+## 1.1.0-alpha.0
+
+### Minor Changes
+
+- Expose token usage from embedding operations ([#12556](https://github.com/mastra-ai/mastra/pull/12556))
+  - `saveMessages` now returns `usage: { tokens: number }` with aggregated token count from all embeddings
+  - `recall` now returns `usage: { tokens: number }` from the vector search query embedding
+  - Updated abstract method signatures in `MastraMemory` to include optional `usage` in return types
+
+  This allows users to track embedding token usage when using the Memory class.
+
+### Patch Changes
+
+- Updated dependencies [[`e6fc281`](https://github.com/mastra-ai/mastra/commit/e6fc281896a3584e9e06465b356a44fe7faade65), [`97be6c8`](https://github.com/mastra-ai/mastra/commit/97be6c8963130fca8a664fcf99d7b3a38e463595), [`5fe1fe0`](https://github.com/mastra-ai/mastra/commit/5fe1fe0109faf2c87db34b725d8a4571a594f80e), [`f6673b8`](https://github.com/mastra-ai/mastra/commit/f6673b893b65b7d273ad25ead42e990704cc1e17), [`cd6be8a`](https://github.com/mastra-ai/mastra/commit/cd6be8ad32741cd41cabf508355bb31b71e8a5bd), [`9eb4e8e`](https://github.com/mastra-ai/mastra/commit/9eb4e8e39efbdcfff7a40ff2ce07ce2714c65fa8), [`aa37c84`](https://github.com/mastra-ai/mastra/commit/aa37c84d29b7db68c72517337932ef486c316275), [`47eba72`](https://github.com/mastra-ai/mastra/commit/47eba72f0397d0d14fbe324b97940c3d55e5a525)]:
+  - @mastra/core@1.2.0-alpha.0
+
 ## 1.0.1
 
 ### Patch Changes

package/dist/docs/README.md

@@ -33,4 +33,4 @@ docs/
 ## Version
 
 Package: @mastra/memory
-Version: 1.0.1
+Version: 1.1.0-alpha.0

package/dist/docs/SKILL.md

@@ -5,7 +5,7 @@ description: Documentation for @mastra/memory. Includes links to type definition
 
 # @mastra/memory Documentation
 
-> **Version**: 1.0.1
+> **Version**: 1.1.0-alpha.0
 > **Package**: @mastra/memory
 
 ## Quick Navigation

package/dist/docs/SOURCE_MAP.json

@@ -1,5 +1,5 @@
 {
-  "version": "1.0.1",
+  "version": "1.1.0-alpha.0",
   "package": "@mastra/memory",
   "exports": {
     "extractWorkingMemoryContent": {

package/dist/index.cjs

@@ -14671,8 +14671,11 @@ var Memory = class extends memory.MastraMemory {
         `Memory error: Resource-scoped semantic recall is enabled but no resourceId was provided. Either provide a resourceId or explicitly set semanticRecall.scope to 'thread'.`
       );
     }
+    let usage;
     if (config?.semanticRecall && vectorSearchString && this.vector) {
-      const { embeddings, dimension } = await this.embedMessageContent(vectorSearchString);
+      const result = await this.embedMessageContent(vectorSearchString);
+      usage = result.usage;
+      const { embeddings, dimension } = result;
       const { indexName } = await this.createEmbeddingIndex(dimension, config);
       await Promise.all(
         embeddings.map(async (embedding) => {
@@ -14716,7 +14719,7 @@ var Memory = class extends memory.MastraMemory {
     const rawMessages = shouldGetNewestAndReverse ? paginatedResult.messages.reverse() : paginatedResult.messages;
     const list = new agent.MessageList({ threadId, resourceId }).add(rawMessages, "memory");
     const messages = list.get.all.db();
-    return { messages };
+    return { messages, usage };
   }
   async getThreadById({ threadId }) {
     const memoryStore = await this.getMemoryStore();
@@ -14963,10 +14966,11 @@ ${workingMemory}`;
       ...this.embedderOptions || {}
     });
     if (isFastEmbed && !this.firstEmbed) this.firstEmbed = promise;
-    const { embeddings } = await promise;
+    const { embeddings, usage } = await promise;
     const result = {
       embeddings,
       chunks,
+      usage,
       dimension: embeddings[0]?.length
     };
     this.embeddingCache.set(key, result);
@@ -14987,6 +14991,7 @@ ${workingMemory}`;
     const result = await memoryStore.saveMessages({
       messages: dbMessages
     });
+    let totalTokens = 0;
     if (this.vector && config.semanticRecall) {
       const embeddingData = [];
       let dimension;
@@ -15002,6 +15007,9 @@ ${workingMemory}`;
           if (!textForEmbedding) return;
           const result2 = await this.embedMessageContent(textForEmbedding);
           dimension = result2.dimension;
+          if (result2.usage?.tokens) {
+            totalTokens += result2.usage.tokens;
+          }
           embeddingData.push({
             embeddings: result2.embeddings,
             metadata: result2.chunks.map(() => ({
@@ -15030,7 +15038,7 @@ ${workingMemory}`;
         });
       }
     }
-    return result;
+    return { ...result, usage: totalTokens > 0 ? { tokens: totalTokens } : void 0 };
   }
   updateMessageToHideWorkingMemoryV2(message) {
     const newMessage = { ...message };