feat: Support Together AI in OpenAIEmbeddings wrapper

docs/_sidebar.md

@@ -74,6 +74,7 @@
         - [GCP Vertex AI](/modules/retrieval/text_embedding/integrations/gcp_vertex_ai.md)
         - [Ollama](/modules/retrieval/text_embedding/integrations/ollama.md)
         - [Mistral AI](/modules/retrieval/text_embedding/integrations/mistralai.md)
+        - [Together AI](/modules/retrieval/text_embedding/integrations/together_ai.md)
         - [Prem App](/modules/retrieval/text_embedding/integrations/prem.md)
     - [Vector stores](/modules/retrieval/vector_stores/vector_stores.md)
       - Integrations

docs/modules/retrieval/text_embedding/integrations/openai.md

@@ -1,11 +1,22 @@
-# OpenAI
+# OpenAIEmbeddings
 
-Let's load the OpenAI Embedding class.
+You can use the `OpenAIEmbeddings` wrapper to consume OpenAI embedding models.
 
 ```dart
-final openaiApiKey = Platform.environment['OPENAI_API_KEY'];
-final embeddings = OpenAIEmbeddings(apiKey: openaiApiKey);
-const text = 'This is a test document.';
-final res = await embeddings.embedQuery(text);
-final res = await embeddings.embedDocuments([text]);
+final openAiApiKey = Platform.environment['OPENAI_API_KEY'];
+final embeddings = OpenAIEmbeddings(apiKey: openAiApiKey);
+
+// Embedding a document
+const doc = Document(pageContent: 'This is a test document.');
+final res1 = await embeddings.embedDocuments([doc]);
+print(res1);
+// [[-0.003105443, 0.011136302, -0.0040295827, -0.011749065, ...]]
+
+// Embedding a retrieval query
+const text = 'This is a test query.';
+final res2 = await embeddings.embedQuery(text);
+print(res2);
+// [-0.005047946, 0.0050882488, -0.0051957234, -0.019143905, ...]
+
+embeddings.close();
 ```

docs/modules/retrieval/text_embedding/integrations/together_ai.md

@@ -0,0 +1,30 @@
+# Together AI Embeddings
+
+[Together AI](https://www.together.ai/) offers several leading [embedding models](https://docs.together.ai/docs/embedding-models#embedding-models) through its OpenAI compatible API.
+
+You can consume Together AI API using the `OpenAIEmbeddings` wrapper in the same way you would use the OpenAI API.
+
+The only difference is that you need to change the base URL to `https://api.together.xyz/v1`:
+
+```dart
+final togetherAiApiKey = Platform.environment['TOGETHER_AI_API_KEY'];
+final embeddings = OpenAIEmbeddings(
+  apiKey: togetherAiApiKey,
+  baseUrl: 'https://api.together.xyz/v1',
+  model: 'togethercomputer/m2-bert-80M-32k-retrieval',
+);
+
+// Embedding a document
+const doc = Document(pageContent: 'This is a test document.');
+final res1 = await embeddings.embedDocuments([doc]);
+print(res1);
+// [[-0.038838703, 0.0580902, 0.022614542, 0.0078403875, ...]]
+
+// Embedding a retrieval query
+const text = 'This is a test query.';
+final res2 = await embeddings.embedQuery(text);
+print(res2);
+// [-0.019722218, 0.04656633, -0.0074559706, 0.005712764, ...]
+
+embeddings.close();
+```

examples/docs_examples/bin/modules/retrieval/text_embedding/integrations/openai.dart

@@ -0,0 +1,24 @@
+// ignore_for_file: avoid_print
+import 'dart:io';
+
+import 'package:langchain/langchain.dart';
+import 'package:langchain_openai/langchain_openai.dart';
+
+void main(final List<String> arguments) async {
+  final openAiApiKey = Platform.environment['OPENAI_API_KEY'];
+  final embeddings = OpenAIEmbeddings(apiKey: openAiApiKey);
+
+  // Embedding a document
+  const doc = Document(pageContent: 'This is a test document.');
+  final res1 = await embeddings.embedDocuments([doc]);
+  print(res1);
+  // [[-0.003105443, 0.011136302, -0.0040295827, -0.011749065, ...]]
+
+  // Embedding a retrieval query
+  const text = 'This is a test query.';
+  final res2 = await embeddings.embedQuery(text);
+  print(res2);
+  // [-0.005047946, 0.0050882488, -0.0051957234, -0.019143905, ...]
+
+  embeddings.close();
+}

examples/docs_examples/bin/modules/retrieval/text_embedding/integrations/together_ai.dart

@@ -0,0 +1,28 @@
+// ignore_for_file: avoid_print
+import 'dart:io';
+
+import 'package:langchain/langchain.dart';
+import 'package:langchain_openai/langchain_openai.dart';
+
+void main(final List<String> arguments) async {
+  final togetherAiApiKey = Platform.environment['TOGETHER_AI_API_KEY'];
+  final embeddings = OpenAIEmbeddings(
+    apiKey: togetherAiApiKey,
+    baseUrl: 'https://api.together.xyz/v1',
+    model: 'togethercomputer/m2-bert-80M-32k-retrieval',
+  );
+
+  // Embedding a document
+  const doc = Document(pageContent: 'This is a test document.');
+  final res1 = await embeddings.embedDocuments([doc]);
+  print(res1);
+  // [[-0.038838703, 0.0580902, 0.022614542, 0.0078403875, ...]]
+
+  // Embedding a retrieval query
+  const text = 'This is a test query.';
+  final res2 = await embeddings.embedQuery(text);
+  print(res2);
+  // [-0.019722218, 0.04656633, -0.0074559706, 0.005712764, ...]
+
+  embeddings.close();
+}

packages/langchain_openai/lib/src/embeddings/openai.dart

@@ -13,6 +13,8 @@ import 'package:openai_dart/openai_dart.dart';
 /// - [Embeddings guide](https://platform.openai.com/docs/guides/embeddings/limitations-risks)
 /// - [Embeddings API docs](https://platform.openai.com/docs/api-reference/embeddings)
 ///
+/// You can also use this wrapper to consume OpenAI-compatible APIs like [Together AI](https://www.together.ai).
+///
 /// ### Authentication
 ///
 /// The OpenAI API uses API keys for authentication. Visit your
@@ -122,7 +124,7 @@ class OpenAIEmbeddings implements Embeddings {
   OpenAIEmbeddings({
     final String? apiKey,
     final String? organization,
-    final String? baseUrl,
+    final String baseUrl = 'https://api.openai.com/v1',
     final Map<String, String>? headers,
     final Map<String, dynamic>? queryParams,
     final http.Client? client,
@@ -199,4 +201,9 @@ class OpenAIEmbeddings implements Embeddings {
     );
     return data.data.first.embeddingVector;
   }
+
+  /// Closes the client and cleans up any resources associated with it.
+  void close() {
+    _client.endSession();
+  }
 }

packages/langchain_openai/test/embeddings/together_ai_embeddings_test.dart

@@ -0,0 +1,43 @@
+@TestOn('vm')
+library; // Uses dart:io
+
+import 'dart:io';
+
+import 'package:langchain_openai/langchain_openai.dart';
+import 'package:test/test.dart';
+
+void main() {
+  group('Together AI Embeddings tests', () {
+    late OpenAIEmbeddings embeddings;
+
+    setUp(() async {
+      embeddings = OpenAIEmbeddings(
+        apiKey: Platform.environment['TOGETHER_AI_API_KEY'],
+        baseUrl: 'https://api.together.xyz/v1',
+      );
+    });
+
+    tearDown(() {
+      embeddings.close();
+    });
+
+    test('Test AI Embeddings models', () async {
+      final models = [
+        'togethercomputer/m2-bert-80M-2k-retrieval',
+        'togethercomputer/m2-bert-80M-8k-retrieval',
+        'togethercomputer/m2-bert-80M-32k-retrieval',
+        'WhereIsAI/UAE-Large-V1',
+        'BAAI/bge-large-en-v1.5',
+        'BAAI/bge-base-en-v1.5',
+        'sentence-transformers/msmarco-bert-base-dot-v5',
+        'bert-base-uncased',
+      ];
+      for (final model in models) {
+        embeddings.model = model;
+        final res = await embeddings.embedQuery('Hello world');
+        expect(res.length, greaterThan(0));
+        await Future<void>.delayed(const Duration(seconds: 1)); // Rate limit
+      }
+    });
+  });
+}